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 |
|---|---|---|---|---|---|
Activation of the mammalian Notch receptor after ligand binding relies on a succession of events including metalloprotease-cleavage , endocytosis , monoubiquitination , and eventually processing by the gamma-secretase , giving rise to a soluble , transcriptionally active molecule . The Notch1 receptor was proposed to be monoubiquitinated before its gamma-secretase cleavage; the targeted lysine has been localized to its submembrane domain . Investigating how this step might be regulated by a deubiquitinase ( DUB ) activity will provide new insight for understanding Notch receptor activation and downstream signaling . An immunofluorescence-based screening of an shRNA library allowed us to identify eIF3f , previously known as one of the subunits of the translation initiation factor eIF3 , as a DUB targeting the activated Notch receptor . We show that eIF3f has an intrinsic DUB activity . Knocking down eIF3f leads to an accumulation of monoubiquitinated forms of activated Notch , an effect counteracted by murine WT eIF3f but not by a catalytically inactive mutant . We also show that eIF3f is recruited to activated Notch on endocytic vesicles by the putative E3 ubiquitin ligase Deltex1 , which serves as a bridging factor . Finally , catalytically inactive forms of eIF3f as well as shRNAs targeting eIF3f repress Notch activation in a coculture assay , showing that eIF3f is a new positive regulator of the Notch pathway . Our results support two new and provocative conclusions: ( 1 ) The activated form of Notch needs to be deubiquitinated before being processed by the gamma-secretase activity and entering the nucleus , where it fulfills its transcriptional function . ( 2 ) The enzyme accounting for this deubiquitinase activity is eIF3f , known so far as a translation initiation factor . These data improve our knowledge of Notch signaling but also open new avenues of research on the Zomes family and the translation initiation factors .
Notch signaling relies on two consecutive cleavages of the receptor after binding of its ligand expressed by a neighboring cell . These two processing steps successively performed by a protease of the ADAM family and by the γ-secretase complex can occur only if the activated receptors on one side , the ligands on the other side , undergo post-translational modifications and trafficking . Some of these complex events begin to be elucidated [1]–[7] . They essentially depend on ubiquitination events affecting the ligand and/or the receptor , and probably regulating sorting and trafficking of the activated versus non-activated molecules . Eventually , after proteolytic release the intracellular portion of Notch ( hereafter named NIC ) enters the nucleus , where it functions as a transcriptional co-activator of Notch target genes . In mammals , the Notch1 receptor was proposed to be monoubiquitinated before its γ-secretase cleavage; the targeted lysine has been localized to its submembrane domain [8] . Investigating how this monoubiquitination is regulated may be crucial for understanding Notch receptor activation and downstream signaling . Ubiquitination is a reversible process , and deubiquitinating enzymes ( DUBs ) remove the ubiquitin moieties from ubiquitinated substrates , thus allowing a tight control of these modifications [9] . A potential deubiquitination step could either affect NIC production by γ-secretase , NIC release from the endocytic vesicles , NIC entry into the nucleus , NIC interaction with its transcriptional cofactors , NIC transcriptional activity , or NIC stability . With the aim of identifying a DUB involved in Notch signaling , we established a screening strategy using shRNA vectors targeting the putative and known DUBs of the human genome [10] . Here , we report the identification of eIF3f as a DUB targeting the activated Notch receptor and positively regulating Notch signaling . eIF3f ( for eukaryotic translation initiation factor 3 subunit f ) is one of the 13 subunits ( named eIF3a-m ) of the translation initiation factor eIF3 . eIF3 stimulates many steps of the translation initiation pathway , including assembly of the eIF2-GTP/met-tRNA complex to the 40S ribosome to form the 43S preinitiation complex ( PIC ) , mRNA recruitment to the 43S PIC complex , impairment of the 40S ribosome to join the 60S prematurely , and scanning the mRNA for AUG recognition [11] . eIF3 has no known enzymatic activity so far , but it has an intriguing degree of homology with two other complexes whose functions appear unrelated: the COP9 signalosome and the 19S proteasome lid . All three complexes , forming the Zomes family , consist of subunits with either PCI ( Proteasome-COP9 signalosome-initiation factor 3 domain ) or MPN ( for MPR1-PAD1-N-terminal domain ) signature domains and share a common 6PCI +2 MPN domain stoechiometry . The mammalian eIF3 also has an additional five non-PCI-MPN subunits [12] . The MPN domain of CSN5 ( COP9 Signalosome subunit 5 ) harbors a metalloprotease motif referred to as the Jab/MPN domain-associated metallopeptidase ( JAMM ) motif and regulates the activity of E3 ubiquitin ligases by deneddylation of the cullin component . On the other hand , a JAMM-containing subunit associated with the 19S proteasome lid ( Rpn11 , [13] ) also harbors DUB activity , accounting for substrates deubiquitination before they enter the proteasome channel . Interestingly , eIF3f contains a JAMM domain , making it a putative DUB [14] , [15] . We show here that it harbors a DUB activity acting on Notch signaling .
In order to identify DUBs involved in the Notch signaling pathway , we set up an immunofluorescence screen . We used Notch ΔE [16] , a mutant form of Notch deleted of most of its extracellular domain . ΔE mimics the ADAM-cleavage product and therefore represents a constitutively active form , which is monoubiquitinated and endocytosed [8] before being cleaved by γ-secretase to liberate NIC [17] . We made use of the V1744 antibody to monitor the production and the localization of NIC , whereas anti-myc antibody detected all Notch products . U2OS cells were co-transfected with vectors encoding ΔE and with each individual pool of a shRNA library targeting the 91 known or putative DUBs encoded by the human genome ( called shDUB library; see Table 1 ) [10] . With ΔE alone ( Figure 1A , Panel A ) , we observed a membrane and endomembrane-localized myc labeling corresponding to the unprocessed Notch ΔE , but also a nuclear myc labeling ( A3 ) corresponding to NIC , co-stained with the V1744 antibody ( A4 ) . The same pattern was observed with almost every shRNA pool of the library ( exemplified in Panels B1–4 ) , whereas four pools abolished ΔE expression ( those targeting eIF3h , PRP8 and USP54 , and Pool 2 , see below ) . However , with an shRNA pool targeting eIF3f ( panels C ) , we observed a partial extra-nuclear V1744 labeling ( C4 ) , as well as an increase in the average proportion of extranuclear/nuclear myc staining ( see Figure S1 , Dose 1 ) . This suggests that NIC production and localization was affected when eIF3f was knocked down . The shRNA library actually contained two pools ( each containing four shRNA-encoding plasmids ) targeting eIF3f: Pool 1 , whose effects are shown in Figure 1A , and Pool 2 , in the presence of which no ΔE-positive signal could be detected by immunofluorescence ( unpublished data ) . We isolated the eight shRNAs from these pools ( six different shRNAs in total ) and tested them individually for their efficiency in knocking down eIF3f: Pool 1 ( P1 ) contains shRNA #1 to #4 , Pool 2 ( P2 ) contains shRNA #1 , #2 , #5 , and #6 . HEK293T cells were transfected with each of these shRNAs and the levels of endogenous or overexpressed human HA-tagged eIF3f were analyzed by Western blotting . Anti-α-Tubulin was used as a loading control ( Figure 1B , bottom panel ) , since this protein is very stable and reflects the total protein content in transfected and non-transfected cells . shRNAs #1 and #2 significantly affected the level of endogenous and transfected eIF3f ( Figure 1B , Lanes B and C and quantification under the lanes ) , similar to the Pool 1 ( P1 , Lane I ) . The additional shRNAs from Pool 2 ( #5 and #6 in Lanes J and K ) and Pool 2 itself ( P2 , Lane L ) almost completely abolished eIF3f expression in this assay . The effect of each of these shRNAs on eIF3f protein level was correlated with its ability to inhibit ΔE expression in immunofluorescence experiments and probably reflects eIF3f requirement for translation ( Figure S1 ) . Indeed shRNAs #3 and #4 , which exhibited a minor effect on the expression of the eIF3f protein ( Figure 1B , Lanes D–G ) , were the most efficient in inducing extranuclear V1744 labeling in a dose-dependent manner ( Figure 1C ) . Nevertheless , further increasing the dose of these two shRNAs separately or co-transfecting them resulted in an effect on eIF3f expression ( Figure 1B , Lanes E , G , and H ) and an extinction of the ΔE signal in immunofluorescence ( Figure S1 and unpublished data ) . Since the shRNAs that affect Notch signaling target different sequences of eI3Ff , and since this effect is complemented by transfection of wt eIF3f ( see below ) , we can exclude an off-target effect of these shRNAs . Taken together , these results suggest that a specific but mild knockdown of eIF3f is responsible for the altered Notch localization observed in Figure 1 . We next tested the effect of eIF3f on the ubiquitination of activated Notch . HEK293T cells were transfected with vectors encoding ΔE and a 6xHis-tagged Ubiquitin . We added increasing amounts of shRNA #3 , without reaching the doses used in Figure 1B , Lane E , in the presence or not of a murine form of eIF3f ( meIF3f ) . The murine eIF3f cDNA exhibits a three base pair change in the sequence targeted by shRNA #3 and is therefore refractory to its effect . Proteins were extracted in denaturing conditions and ubiquitinated proteins were purified on Nickel-charged beads . Finally , whole cell extracts and ubiquitinated products were analyzed by Western blot to quantify the levels of ubiquitinated Notch ( Figure 2A ) . Transfection of shRNA #3 led to a dose-dependent accumulation of monoubiquitinated ΔE ( ΔEUb ) and of monoubiquitinated NIC ( NICUb ) ( Lanes C , D compared to B ) , whereas the levels of Notch in the extracts remained stable ( Lanes G to J ) . The same effect was observed using shRNA #4 ( unpublished data ) . Interestingly , overexpression of meIF3f abolished the effect of shRNA #3 and accumulation of ubiquitinated Notch could no longer be seen ( compare Lanes E to D ) . The ubiquitination levels of other forms of Notch were also analyzed in the presence or absence of shRNA #3 ( Figure 2B ) : the membrane-anchored ΔE-LLFF ( Lanes D–F ) , which undergoes the same processes as ΔE ( i . e . , monoubiquitination and endocytosis ) but which is mutated in the γ-secretase cleavage site and consequently does not generate NIC [18]; the nuclear Notch ( NIC , Lanes G–I ) and the non-activated full-length Notch ( FL , Lanes J–L ) . The only forms whose ubiquitination was affected by shRNA #3 and/or meIF3f were those corresponding to activated and still membrane-anchored forms of Notch , namely ΔE and ΔE-LLFF . We also tested as controls NDFIP2 ( a transmembrane protein located in the endosomal compartment [19] ) and Deltex1 ( an E3 ubiquitin ligase genetically identified as involved in Notch signaling ) . These two proteins showed no change in ubiquitination in the presence of shRNA #3 or overexpressed meIF3f ( Figure S2 and unpublished data ) . These results suggest that eIF3f is specifically involved in an ubiquitination/deubiquitination process targeting activated Notch . The eIF3f protein contains an MPN domain , also found in some DUBs of the JAMM family , such as AMSH , Rpn11 , and CSN5 [20]–[23] . However , the position of the amino acids constituting the catalytic site of these DUBs is not strictly conserved in eIF3f , although histidines and acidic residues can still be found and can possibly form a signature of metalloprotease ( HEX2HX2GX2H ) . We mutated six amino acids of eIF3f , including two histidines and two acidic residues of the putative catalytic site . We then tested in parallel WT meIF3f and this putative catalytic mutant ( meIF3f Mut ) ( Figure 2C ) . Whereas WT meIF3f was able to prevent shRNA #3–induced NICUb accumulation in a dose-dependent manner ( Lanes D , E compared to C ) , meIF3f Mut was not , although it was expressed at similar levels ( Lanes D to G ) . This shows that meIF3f indeed complements the effect of shRNA #3 and that an intact MPN domain of eIF3f is necessary for its effect on monoubiquitination of activated Notch . Taken together , these results strongly suggest that eIF3f could act as a DUB targeting activated Notch . In order to test whether eIF3f exhibits a deubiquitinase activity , we first used a functional in bacteria assay using GFP fused to ubiquitin ( Ub-GFP ) [24] . The peptide bond between ubiquitin and GFP can be cleaved by ubiquitin- or UbL-deconjugase activities , which are absent from the bacterial genome . Bl21 bacteria were transfected with plasmids encoding GST alone , GST-fused to human WT eIF3f or WT MPN domain , or His-tagged murine eIF3f WT or mutant MPN domain ( Mut ) , together with a vector encoding Ub-GFP fused to an S-Tag at its C-terminus ( Ub-GFP-S-Tag ) ( Figure 3A ) . As control DUB , we used BPLF1 WT ( an EBV DUB of the cysteine-protease family [24] ) and its catalytically inactive form BPLF1 CM . After induction of protein expression , the bacteria were lysed and cleavage of Ub-GFP was assessed in Western blot by the appearance of free GFP-S-Tag using anti-S-Tag antibody ( Figure 3A , upper panel ) . Anti-GST or anti-His antibodies were used to verify protein expression ( Figure 3A , bottom panel ) . As expected , we observed released GFP with BPLF1 WT but not its CM mutant ( compare Lanes A–C ) . Interestingly , we also detected protease activity with heIF3f WT , heIF3f MPN WT , and meIF3f MPN WT ( Lanes D , E , and F , respectively ) but not with the mutant meIF3f MPN Mut ( Lane G ) . To further investigate whether eIF3f could act as an ubiquitin-specific protease , we performed an in vitro assay using Ubiquitin Vinyl Sulfone ( Ub-VS ) , a functional probe that covalently binds the active site of DUBs and consequently inhibits DUB catalytic activity ( Figure 3B ) . HeLa cells were transfected with vectors encoding HA-tagged meIF3f , either WT or Mut . Flag-tagged BPLF1 WT and CM were used as controls . A portion of the corresponding whole cell extracts was incubated in vitro with Ub-VS . Total extracts incubated or not with Ub-VS were finally analyzed by Western blot using anti-HA and anti-Flag antibodies to monitor the covalent binding of Ub-VS to the DUBs ( Figure 3B , Lanes A to H ) and to check protein expression ( Lanes I to N ) , respectively . We observed a partial upshift corresponding to the size of the Ub-VS probe with meIF3f WT ( Lane F compared to E ) and with BPLF1 WT ( Lane B compared to A ) , indicating that both are able to bind Ub-VS . In contrast , we did not observe any shift with meIF3f Mut ( Lane H compared to G ) nor BPLF1 CM ( Lane D compared to C ) , showing that they are indeed unable to bind Ub-VS . We obtained the same results when the DUBs were expressed in bacteria ( unpublished data ) . Taken together , these results show that eIF3f exhibits a deubiquitinase activity , carried by its MPN domain . Moreover it confirms that the mutant we generated ( Mut ) by replacing six amino acids of the metalloprotease-like sequence of eIF3f is a catalytically inactive form of eIF3f . In order to identify the site of action and the target of eIF3f in the Notch activation cascade , we first overexpressed ΔE and murine eIF3f in U2OS cells . We observed by immunofluorescence a low frequency of vesicular colocalization of the two proteins ( Figure 4 , Panels A and B ) . As eIF3f might require a cofactor or target an intermediate protein to regulate Notch ubiquitination , we tested several components of the Notch signaling pathway that could be associated with trafficking . Some DUBs are known to be recruited to their substrate indirectly via an E3 ubiquitin ligase [25] , [26] , so we particularly focused on the E3 ubiquitin ligases of the Notch pathway known to be associated with the endocytic machinery: Itch/AIP4 and Deltex1 ( hereafter designated as DTX ) ( [27] and therein ) . In contrast to Itch/AIP4 , DTX significantly colocalized with eIF3f when both proteins were coexpressed ( Figure 4 , Panel C ) . In addition , the presence of DTX strikingly increased the colocalization of ΔE and eIF3f , the three proteins being associated to the same vesicles ( Figure 4 , Panel D ) . These results suggest that DTX could recruit eIF3f to activated Notch . To verify this hypothesis , we performed co-immunoprecipitation experiments in HEK293T cells transfected with vectors encoding VSV-tagged DTX , Flag-Itch/AIP4 , and HA- or Flag-tagged forms of eIF3f ( schematized in Figure 5A ) . We pulled down eIF3f and analyzed the whole cell extracts and the immunoprecipitates by Western blot ( Figure 5B , 5C and Figure S3 ) . DTX co-immunoprecipitated with eIF3f WT and ( 1–192 ) , but also with eIF3f ( 188–361 ) and ( 91–361 ) ( compare Lanes B , C to D , E in Figure 5B and 5C ) . No co-immunoprecipitation was detected with Itch/AIP4 ( Figure S3 ) . As a control , the endogenous eIF3a , another subunit of the eIF3 complex , only co-immunoprecipitated with eIF3f WT and ( 91–361 ) ( Panels B , C ) . This observation suggests that DTX is able to physically interact with eIF3f and that the domain ( s ) necessary for this interaction is different from the domain necessary for eIF3f to incorporate the eIF3 complex . We then performed the reverse experiment by pulling down VSV-tagged DTX and could detect the various forms of eIF3f coimmunoprecipitating with DTX ( Figure 5D ) : WT or Mut eIF3f ( Lanes B , C ) as well as the deletion mutants ( Lanes D–F ) . To confirm these results under conditions where the proteins are not overexpressed , we established by retroviral transduction murine cell lines expressing low amounts of VSV-DTX and S-tagged forms of WT eIF3f or of a mutant of the active site ( HDI to AAA ) . eIF3f or DTX was immunoprecipitated first and their association was confirmed in both cases ( Figure 5E ) . We then performed co-immunoprecipitations in HEK293T in the presence of Notch ΔE ( Figure 6A ) . While DTX co-immunoprecipitated with WT eIF3f but also with meIF3f Mut ( Figure 6A , Lanes B–E ) , ΔE did not ( Lanes I and J ) , unless DTX was cotransfected ( Lanes D and E ) . This strongly suggests that a tripartite interaction occurs between activated Notch , DTX , and eIF3f , DTX being required for the Notch-eIF3f interaction . In addition , we verified that neither DTX nor ΔE could be co-immunoprecipitated with eIF3f from cell extracts that were transfected separately and mixed ( unpublished data ) . We then repeated these experiments using other forms of Notch: ΔE-LLFF , NIC , or the non-activated full-length Notch ( FL ) ( Figure 6B ) . Only two forms could co-immunoprecipitate with eIF3f in the presence of DTX: ΔE and ΔE-LLFF ( Lanes C and D ) . No signal was detected with FL and NIC ( Lanes E and F ) even with a longer exposure . It is of note that NIC produced from ΔE ( Lanes C , G ) and detected by the V1744 antibody was not co-immunoprecipitated with eIF3f either . These results show that the tripartite interaction between Notch , DTX , and eIF3f occurs preferentially with activated , membrane-associated , but γ-secretase unprocessed , forms of Notch . Our results ( Figures 2 and 6 ) suggest that the target of eIF3f is preferentially a γ-secretase unprocessed form of activated Notch . The fact that we could detect ubiquitinated NIC in the presence of shRNAs #3 and #4 ( Figure 2A ) might thus appear paradoxical , however it might be the consequence of γ-secretase cleavage of non-deubiquitinated Notch ΔE . We also observed that shRNAs #3 and #4 led to a partially extranuclear localization of NIC ( Figure 1C ) . This suggests that NIC carrying residual monoubiquitination could be impaired in nuclear translocation . In order to test this possibility , we tried to mimic NICUb by attaching an ubiquitin to the N-terminus of NIC ( UBIC construct ) . Given that ΔE , and consequently any putative non-deubiquitinated NIC , has been shown to be monoubiquitinated on K1749 , we used a K1749R mutant of NIC ( NIC ( KR ) ) on which no ubiquitin can be conjugated . In addition , the attached ubiquitin was mutated on the internal Lysine residues 29 and 48 to impair polyubiquitination and also on the 2 C-terminal glycines to prevent proteolysis by a DUB enzyme [28] . U2OS cells were transfected with NIC , NIC ( KR ) , or the new construct UBIC . We observed by immunofluorescence that NIC or NIC ( KR ) were mostly nuclear ( 96% in average ) , whereas UBIC was partially retained in the cytoplasm ( 18% see Figure S4 , Panel A ) . In addition , we monitored the transcriptional activity of these three forms by cotransfecting U2OS cells with increasing doses of expression vectors together with a Notch-reporter gene ( CSL-luciferase , [29] , [30] ) and an internal control reporter ( pRL-TK ) . As shown in Figure S4 , Panel B , UBIC is significantly less active than NIC or NIC ( KR ) , although it is expressed at a comparable level ( see Western blot in Figure S4 , Panel B bottom ) . One possibility to explain the localization and the drop in transcriptional activity of UBIC is that the ubiquitin moiety partially prevents access to the NLS of NIC . In order to test whether eIF3f could act on Notch signaling under more physiological conditions , we performed a coculture assay using a CSL reporter strategy . U2OS cells stably expressing Notch FL were transfected with a CSL-Luciferase Notch reporter and increasing doses of meIF3f WT , meIF3f Mut , or meIF3f ( 188–361 ) . pRL-TK vector , encoding Renilla luciferase under the control of the Notch-insensitive thymidine kinase promoter , was also cotransfected as an internal control . These cells were then cocultured with OP9 cells stably expressing or not the Notch ligand Delta-like1 ( Dll1 ) [31] , and relative luciferase activity was finally determined by normalizing CSL-firefly luciferase with renilla luciferase . In parallel cell extracts from the same transfections were analyzed by Western blot ( Figure 7A , bottom ) . In the presence of Dll1 , the relative luciferase activity of the CSL reporter gene increased 20-fold ( Figure 7A , Lanes A , B ) , showing that Notch was indeed activated by Dll1 . While the presence of meIF3F WT did not modify Dll1-dependent Notch activation ( Lanes C to F ) , meIF3F Mut and meIF3f ( 188–361 ) repressed it in a dose-dependent manner , respectively , reaching 35% and 54% of reduction ( Lanes G–J and K–N , respectively ) . Thus , meIF3f Mut and meIF3f ( 188–361 ) inhibit Notch transcriptional activity and behave as dominant-negative forms of eIF3f , in accordance with the fact that both lack an intact catalytic site but are able to bind DTX and Notch . We also tested the effect of shRNAs targeting endogenous eIF3f in this assay . As represented in Figure 7B , Dll1-induced Notch stimulation was inhibited in the presence of increasing doses of either P2 ( Lanes C–E ) or shRNA #1 ( Lanes F–H ) , although Notch1 overall level remained constant ( see bottom of Figure 7B ) . In contrast , a shRNA pool targeting AMSH , another DUB of the JAMM family , had no effect on Notch activation ( Lanes I–K ) . Furthermore , the same shRNAs have no effect on NIC-mediated transcriptional activation ( Figure S5 ) , excluding any effect of the shRNAs on Notch-associated transcription factors . Therefore , inhibition of eIF3f DUB activity impairs the production of a transcriptionally active NIC in a dose-dependent manner . All together , our results indicate that eIF3f , after being recruited to vesicular Notch via DTX , could act as a DUB on a monoubiquitinated but not γ-secretase processed form of Notch , thus positively modulating Notch signaling activity .
With the aim of identifying a DUB that could regulate Notch activation , we reached two new and provocative conclusions: the activated form of Notch needs to be deubiquitinated to enter the nucleus and fulfill its transcriptional function , and the DUB accounting for this activity is eIF3f , known so far as a translation initiation factor . When reconstituting a Notch activation system by co-culturing Notch1 receptor- and Dll1 ligand-expressing cells , we observed that Notch-dependent transcriptional activation of a reporter gene was specifically affected when expressing mutant forms of eIF3f where the active site was mutated or when eIF3f was partially knocked down . Therefore eIF3f acts as a positive regulator of Notch signaling , independently of its function in translation ( see below ) . We have also demonstrated that a monoubiquitinated form of Notch ΔE was stabilized when eIF3f was slightly knocked down . This effect was accompanied by the appearance of a mono-ubiquitinated form of NIC , which is probably excluded from the nucleus . Therefore we conclude that Notch deubiquitination is necessary for its full activity in the nucleus . We have not formally identified the form of Notch that is the substrate of the DUB activity . We cannot exclude that it is a monoubiquitinated non-nuclear NIC , resulting from γ-secretase cleavage of activated Notch . Such a form has never been detected before , in contrast to the nuclear polyubiquitinated NIC , which appears subsequently under the action of the E3 ubiquitin ligase Sel10 [32] , [33] . An artificial construct mimicking to a certain extent the monoubiquitinated NIC ( UBIC ) was partially retained in the cytoplasm and as a consequence was transcriptionally less active than NIC . This suggests that the ubiquitin moiety could mask the proximal NLS [34] and impair interaction with importins . Nevertheless three observations argue against the possibility of NIC being the natural substrate of eIF3f: first , NIC was not obviously associated with endosomes when it was excluded from the nucleus in the presence of eIF3f shRNA . As eIF3f colocalizes with Notch to endosomal structures , its target should be retained to these structures ( Figure 4 ) . Second , NIC produced from ΔE was not detected as coimmunoprecipitating neither with eIF3f nor with its mutant in the presence of DTX , which would have probably been the case if it were the DUB substrate ( Figure 6 ) . And third , the effect of shRNAs targeting eIF3f were only detected on activated but still membrane-anchored forms of Notch , including one unable to produce NIC ( ΔE-LLFF , see Figure 2 ) . Therefore we favor the hypothesis that the eIF3f substrate is the activated , monoubiquitinated but still membrane-anchored Notch ( mimicked by ΔE in transfection experiments , see [8] ) . The presence of monoubiquitinated NIC when eIF3f activity is impaired would be due to a partial activity of γ-secretase on non-deubiquitinated ΔE . Finally , our results show that eIF3f is recruited to Deltex1 and Notch ΔE-containing vesicles in transfected cells . They can thus form a tripartite complex where DTX serves as a bridging factor between Notch and eIF3f . Deltex1 belongs to the RING family of E3 ubiquitin ligase , however its target in Notch signaling remains to be determined [35] . It localizes to the endocytic pathway [27] , interacts with Notch , and thus could serve as a scaffolding protein enabling Notch trafficking and modifications , and eventually Notch signaling . Our results suggest that eIF3f , known so far as a component of a translation initiation factor complex , is itself able to act as a DUB during Notch activation . In contrast to a mutant of the active site , a WT form of eIF3f is able to complement the inhibition of activity mediated by eIF3f shRNAs and to restore deubiquitination of Notch ΔE . However we cannot completely rule out the possibility that another DUB , associated with eIF3f , could account for Notch deubiquitination . This is , for example , the case for CSN5 , which , in addition to its own isopeptidase activity , is associated with USP15 , both being required for proper processing of polyubiquitinated substrates bound to p97/VCP [36]–[38] . Such a putative eIF3f-associated DUB , the knockdown of which would not affect protein translation , should have been identified during the screen . Moreover , eIF3h , the other MPN-containing subunit of eIF3 associated with eIF3f [12] , has neither histidines nor acidic residues that could confer a DUB activity . The fact that eIF3f was the only DUB identified argues against the possibility of an associated DUB . eIF3f was not identified in other screens , in particular in one recently performed in Drosophila [39] , probably because a strong extinction of this translation factor had a more severe and broad phenotype than a Notch phenotype , even on external sensory organ development . We actually were able to pick up this factor thanks to the presence of relatively inefficient shRNAs in the library . The human genome encodes 14 JAMM proteins , seven of which have a complete set of the conserved residues for Zn2+ coordination [40] . Among them , six ( AMSH , AMSH-LP , BRCC36 , Rpn11 , MYSM1 , and CSN5 ) have been reported to have isopeptidase activity on ubiquitin or ubiquitin-like proteins . It is of note that the JAMM sequence of eIF3f cannot be aligned with those of the proteins constituting the MPN+ group ( Rpn11 , Csn5 , AMSH; see [15] ) . Nevertheless the four polar and the additional glutamate residue in a more N-terminal region of the domain are still present in eIF3f , although not arranged in the MPN+-defined pattern , and the mutation of some of these amino acids affects eIF3f enzymatic activity and function in Notch signaling . It must be noted that the MPN+ consensus domain has been generated using a small number of proteins and that divergent members may well exist , including eIF3f . Using an in bacteria assay , we have demonstrated that full-length eIF3f or its isolated MPN domain exhibit DUB activity and that the active site indeed involves the amino acids that we had targeted in our inactive mutant . The relatively low activity that we detected in this assay , as compared to the BPLF1 control , might be due to the specificity of this DUB , which is limited by the nature of the P'1 and P'2 residues in the substrate sequence , thus preventing cleavage of non-specific substrates . However we have confirmed by Ub-VS fixation that WT but not mutant eIF3f exhibits DUB activity in mammalian cells . Among the members of the JAMM family , AMSH and AMSH-like seem to be specific for Lys63-linked polyubiquitin chains [41] , while MYSM1 is specific for monoubiquitinated H2A [42] . Sato et al . [43] suggested that the specificity of AMSH family members for Lys63-linked polyubiquitin chains is primarily due to their interaction with the proximal ubiquitin , involving a single domain containing two characteristic insertions that are not conserved in eIF3f . However , DUBs may have a multidomain structure , and some of them associate with other proteins , including E3 ubiquitin ligases [23] , [44] . We show that Notch is only able to interact with eIF3f in the presence of the E3 ubiquitin ligase DTX and that a monoubiquitinated form of Notch is probably the substrate targeted by the DUB . Therefore any reconstituted in vitro system will be difficult to set up . As JAMM DUBs are commonly found in association with large protein complexes , on one hand , and eIF3f belongs to the eIF3 complex , on the other hand , it will be of great interest to determine whether eIF3f can work as a DUB outside of the translation complex or whether the active form of eIF3f is associated with the whole translation initiation complex . If it were the case , beside an analogy and homology in the molecular architecture of the other Zomes members ( Cop9 signalosome and proteasome , [45] ) , eIF3 would also harbor an enzymatically-conserved organization . Beyond acting as a translation initiation factor , eIF3f fulfills other functions , probably independently of the eIF3 complex . For instance , it was shown to inhibit HIV-1 replication [46] by modulating the sequence-specific recognition of the HIV-1 pre-mRNA by the splice factor 9G8 . It was also recently proposed to serve as a scaffold in coordinating mTor and SGK1 actions on skeletal muscle growth [47] and to interact genetically and physically with TRC8 , an E3 ubiquitin ligase of the RING family with several TM domains [48] . On the other hand , there are at least two different types of eIF3 complexes in the cell , localized in different subcellular fractions . One complex lacks eIF3a and eIF3f , while the other consists of eIF3a–c and eIF3f . Phosphorylated eIF3f may predominantly localize to the nucleus and join a complex containing at least b and c during apoptosis . Therefore , the nuclear eIF3 complex is likely to have functions other than translation initiation [49] . In yeast , affinity purification and LC-MS/MS was employed to characterize the eIF3 interactome , which was found to contain 230 proteins [50] . This led to the proposal that eIF3 assembles into a large supercomplex , the translasome , which contains elongation factors , tRNA synthases , 40S and 60S ribosomal proteins , chaperones , and the proteasome . On the other hand , eIF3 also associates with importins-β , a critical event for normal cell growth . These data suggest that translasomes are dynamically localized within the cell , and eIF3 could shuttle between the cytoplasm and the nucleus in a cell cycle-dependent manner . All these data are in agreement with the hypothesis that eIF3f , and maybe part of eIF3 , could have multiple functions in the cells . Our results now add an enzymatic activity to these various properties; it remains to be elucidated whether this activity is necessary to fulfill all eIF3f functions . Recent data suggest that other translation factors , many of which exist in several copies in the genome ( as does eIF3f ) , have “part-time jobs” outside of their usual function in translation [51] . It is the case , for example , for eIF4A , which exhibits an RNA helicase activity in translation initiation and which was shown to act in a translation-independent manner as a negative regulator of Dpp/BMP signaling in drosophila [52] . eIF4A affects Mad protein level and was suggested to act as an E3 ubiquitin ligase . In parallel also to ribosomal proteins , which have been recently shown to function outside of the ribosome and to be recruited in some cases for a function unrelated to the ribosome or its synthesis [53] , we propose that eIF3f would be another example of a housekeeping protein that also plays a specific role in the Notch signaling pathway .
shDUB library was described in [10] ( see also [40] , [54] , [55] ) . Additive shRNAs are listed in Table 1 . The pool targeting AMSH used as a control in Figure 7B is the number 139 of the library . All myc-tagged Notch constructs ( ΔE , ΔE-LLFF , FL , and NIC ) were already described [18] , [56] and were gifts from R . Kopan ( Washington University , St . Louis , MO ) . These Notch constructs are all deleted from aa 2183 of murine Notch1 and fused to a hexameric myc tag at the carboxy terminus . Notch1 retroviral vector encodes a full-length human Notch1 with an HA-epitope tag inserted between EGF repeats 22 and 23 ( gift of J . Aster , Harvard medical school , Boston , USA ) . All WT and deletion mutants of eIF3f ( containing N-terminal tags ) , as well as GST-fused eIF3f constructs , were gifts from S . Leibovitch ( Montpellier , France ) , except the catalytic mutant . meIF3f Mut was generated from HA-tagged WT meIF3f by two successive site-directed mutagenesis using , respectively , the oligonucleotides 5′-GCCACAGGCGCTGCAGCCACAGAACACTCAGTGCTG-3′ and its complementary DNA , and 5′-GACATCACAGCTGCAGCCCTGCTGATCCATGAG-3′ and its complementary DNA ( HDI ( aa181–183 ) mutated in AAA and IHE ( aa190–192 ) mutated in GIL ) . 6xHis-tagged Ubiquitin construct was a gift from M . Treier ( European Molecular Biology Laboratory , Heidelberg , Germany ) . HIS-tagged eIF3f MPN constructs were built by inserting the amino acids 95 to 226 of meIF3f WT or meIF3f Mut in pet28a vector using NdeI and BamHI restriction sites . Retroviral meIF3f vectors were constructed first by adding a S-Tag in the N-terminus end of meIF3f WT or meIF3f mutated in the catalytic domain ( HDI ( aa181–183 ) mutated in AAA ) . Then S-tagged meIF3f constructions were inserted in pMSCVpuro vector using HindIII and ClaI restriction sites . The puromycin resistance gene was replaced by S-Tagged meIF3f , whose expression is by the way under the control of PGK promoter . CSL-Luc was a gift from T . Honjo ( Kyoto University , Japan ) and is referred to as pGa981-6 in [30] . NDFIP2 was a gift from S . Kumar ( Adelaide University , Australia ) and is referred to in [19] . Both BPLF1 constructs were described in [24] . We used anti-myc 9E10 and anti-VSV P5D4 antibodies . Anti-Notch IC and anti-Dll1 antibodies were described , respectively , in [57] and [31] . Other antibodies for Western blot were supplied by Abcam ( polyclonal anti-S-Tag ) , Bethyl ( polyclonal anti-eIF3a ) , BioLegend ( polyclonal anti-eIF3f ) , Cell Signaling ( anti-Notch V1744 ) , Covance ( monoclonal anti-HA; polyclonal anti-HA ) , Invitrogen ( polyclonal anti-GFP ) , Novagen ( monoclonal anti-S-Tag ) , and Sigma ( monoclonal anti-Flag M2; polyclonal anti-Flag; monoclonal anti-β-tubulin; monoclonal anti-α-tubulin , monoclonal anti-β-Actin ) . Secondary antibodies for immunofluorescence were supplied by Molecular Probes ( Alexa Fluor conjugates ) . U20S-FL cell line was established by retroviral transduction . High titers of recombinant HA-tagged Notch FL viruses were obtained 48 h after transfection of the Plat-E ecotropic packaging cell line with retroviral expression plasmids . After retroviral transduction of the U2OS cell line , clonal populations were obtained by limiting dilution . OP9-Dll1 cell line was described in [31] . MEFs stably expressing VSV-DTX and S-Tag-meIF3f were established from the DTX-expressing MEFs [27] by retrotransduction of S-tagged meIF3f vectors . U2OS cells were grown on glass coverslips and were transiently transfected using FuGeneHD transfection reagent ( Roche , Mannheim , Germany ) for 24 h . Cells were fixed with 4% paraformaldehyde and permeabilized with PBS containing 0 . 2% Triton X-100 for 5 min before the incubation with appropriate antibodies . Cell preparations were mounted in Mowiol ( Calbiochem , Merck Biosciences , Darmstadt , Germany ) and images acquired using an AxioImager microscope with ApoTome system with a 63× magnification and AxioVision software ( Carl Zeiss MicroImaging Inc . , Le Pecq , France ) . 293T cells were harvested 24 h after transfection and lysed in 8 M urea , 0 . 1 M NaH2PO4 , 10 mM Tris-Hcl ( pH 8 ) , 1% Triton X-100 , and 20 mM Imidazole at room temperature . His-Ub conjugated proteins were purified on chelating Sepharose beads ( Pharmacia ) , previously charged with Nickel . Ni-bound proteins were washed extensively with the same buffer , then with a pH 6 . 3 buffer , and eluted in Laemmli before Western blot analysis . 293T cells were collected 24 h after transfection , washed in PBS buffer , and lysed in 50 mM Tris-HCl ( pH 7 , 9 ) , 400 mM NaCl , 5 mM MgCl2 , 1% Triton X-100 , supplemented with protease inhibitor cocktail ( Roche ) . Cell extracts were cleared by centrifugation at 14 , 000 rpm for 20 min at 4°C . Immunoprecipitations were performed with the appropriate antibodies in the same buffer . When indicated , the immunoprecipitates were eluted by peptide competition ( 2 mg/mL ) for 1 h at 4°C . Samples were denatured in Laemmli buffer for SDS-PAGE resolution , and immunoblots were performed as described previously [18] . In bacteria functional assay: Bl21 bacteria were transfected with GST-fusions or His-tagged encoding vectors together with Ub-GFP plasmid . The bacteria were selected on agar plates and cultured in LB medium containing Chloramphenicol and Ampicillin or Kanamycin . Exponential bacteria cultures ( OD = 0 . 4 ) were treated with IPTG 0 . 5 mM for 16 h at 23°C . Bacteria were harvested , washed , resuspended in PBS containing 20 mM N-Ethylmaleimide , and sonicated for 30 s on ice . After centrifugation ( 12 , 000 rpm , 10 min ) , clear supernatants were measured for GFP fluorescence and protein concentration . Samples were analyzed by two SDS-PAGE resolutions using for each sample 100 units of fluorescence for Ub-GFP assay and 10 µg extracts for protein expression analysis . Ub-VS assay: HeLa cells were collected 36 h after transfection , washed with PBS , and lysed in 50 mM Tris-Cl pH 7 . 4 , 150 mM NaCl , 1 mM DTT , 1 mM EDTA , 1 mM PMSF , and 0 . 5% NP40 . Protein concentrations were measured . 10 µg of WCE were incubated for 1 h at 37°C with 1 µg of Ub-VS functional probe ( Boston Biochem ) in a 30 µL final volume of labeling buffer ( 50 mM Tris-Cl pH 7 . 4 , 250 mM sucrose , 5 mM MgCl2 , and 1 mM DTT ) . Finally , WCE and Ub-VS-treated samples were analyzed by Western blot . 20 , 000 U2OS-FL cells/cm2 were grown and transiently transfected using FuGeneHD transfection reagent ( Roche , Mannheim , Germany ) . 24 h after transfection , U2OS-FL cells were cocultured with 35 , 000 OP9-Dll1 cells/cm2 . 18 h later , cocultures ( done in triplicates ) were lysed using Passive Lysis buffer ( Promega ) . A fraction of cell lysates were transferred to a white 96-well plate ( Berthold ) . Firefly and Renilla luciferase activities were measured using the luminometer Centro XS ( Berthold ) . For Western blot analysis , part of cocultures were lysed with 8 M urea , 0 . 1 M NaH2PO4 , 10 mM Tris-Hcl ( pH 8 ) , 1% Triton X-100 , and 20 mM Imidazole . | The highly conserved signaling pathway involving the transmembrane receptor Notch is essential for development , and misregulation of this pathway is linked to many diseases . We previously proposed that the Notch1 receptor is monoubiquitinated during its activation . With the aim of identifying a deubiquinating enzyme that could regulate Notch activation , we demonstrated that eIF3f , known previously as part of the multiprotein translation initiation factor eIF3 complex , harbors an enzymatic activity that acts on Notch . The activated form of Notch is able to interact with eIF3f only in the presence of the E3 ubiquitin ligase Deltex , and Notch needs to be deubiquitinated before it can be cleared and its intracellular domain can enter the nucleus and fulfill its transcriptional function . Our results further decipher the molecular mechanisms of Notch signaling activation , showing that ubiquitination and deubiquitination events are required . Additionally , we show that beyond acting as a translation initiation factor , eIF3f fulfills other functions and has an intrinsic enzymatic activity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/cell",
"signaling",
"cell",
"biology/membranes",
"and",
"sorting",
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"biochemistry/transcription",
"and",
"translation"
] | 2010 | The Translation Initiation Factor 3f (eIF3f) Exhibits a Deubiquitinase Activity Regulating Notch Activation |
The survival and persistence of Mycobacterium tuberculosis depends on its capacity to manipulate multiple host defense pathways , including the ability to actively inhibit the death by apoptosis of infected host cells . The genetic basis for this anti-apoptotic activity and its implication for mycobacterial virulence have not been demonstrated or elucidated . Using a novel gain-of-function genetic screen , we demonstrated that inhibition of infection-induced apoptosis of macrophages is controlled by multiple genetic loci in M . tuberculosis . Characterization of one of these loci in detail revealed that the anti-apoptosis activity was attributable to the type I NADH-dehydrogenase of M . tuberculosis , and was mainly due to the subunit of this multicomponent complex encoded by the nuoG gene . Expression of M . tuberculosis nuoG in nonpathogenic mycobacteria endowed them with the ability to inhibit apoptosis of infected human or mouse macrophages , and increased their virulence in a SCID mouse model . Conversely , deletion of nuoG in M . tuberculosis ablated its ability to inhibit macrophage apoptosis and significantly reduced its virulence in mice . These results identify a key component of the genetic basis for an important virulence trait of M . tuberculosis and support a direct causal relationship between virulence of pathogenic mycobacteria and their ability to inhibit macrophage apoptosis .
Tuberculosis is an infectious disease of enormous and increasing global importance . Currently , about one third of all humans are latently infected with its etiologic agent , Mycobacterium tuberculosis ( Mtb ) , and an estimated 2 . 5 million people die of tuberculosis annually [1] . After infection of a mammalian host , Mtb is able to resist innate host defenses sufficiently to increase the local bacterial burden and disseminate throughout the body . With the onset of the adaptive immune response , however , the bacterial numbers are controlled in over 90% of infected individuals . Nevertheless , the host is not able to completely clear the bacterial burden , thus leading to persistence of Mtb within the lungs and other tissues of healthy individuals . These latent infections can be reactivated to generate full-blown disease , a process that is accelerated by immunocompromised states resulting from senescence , malnutrition , and co-infection with HIV , which is a major source of morbidity and mortality associated with the current HIV epidemics in many countries [2–5] . Programmed cell death ( apoptosis ) plays an important role in the innate immune response against pathogens and comprises an evolutionarily conserved defense strategy that extends even into the plant world [6 , 7] . It is therefore essential for persisting intracellular pathogens to have strong anti-apoptosis mechanisms [8–12] . While a few studies have suggested that under some conditions Mtb may induce host cell apoptosis [13–16] , a substantial body of evidence points strongly to the expression of strong anti-apoptotic mechanisms by Mtb and other closely related virulent bacteria . Furthermore , this capacity is not found in avirulent species , suggesting a causal link between virulence and inhibition of macrophage apoptosis [17–19] . This hypothesis is supported by the recent discovery that the genetic predisposition of different inbred mouse strains to mycobacterial infections is linked to the capacity of their macrophages to undergo apoptosis or necrosis upon infection , with the former response imparting a resistant and the latter a susceptible host phenotype [20] . Further confirmation of the findings that Mtb inhibits host cell apoptosis is provided by a number of studies that have addressed its molecular mechanism . The importance of Mtb-induced upregulation of anti-apoptosis genes in infected macrophages for apoptosis inhibition was supported by functional data using either anti-sense oligonucleotides to knock down mcl-1 expression [19] or A1 knock-out mice lacking the anti-apoptosis gene A1 [21 , 22] . These results implicate the intrinsic ( mitochondria-mediated ) apoptosis pathway as a target for Mtb-mediated apoptosis inhibition , because mcl-1 and A1 are both members of the large family of Bcl-2–like proteins that localize prominently to mitochondria . However , this is contradicted by the finding that overexpression of Bcl-2 ( another mitochondrial anti-apoptotic protein ) could not rescue cells from undergoing apoptosis after infection with nonvirulent mycobacteria , thus suggesting that the extrinsic pathway ( death receptor–mediated ) is involved in the infection-induced apoptosis [23] . Consistently , virulent Mtb strains could inhibit FasL-induced apoptosis in Fas-expressing cells [18] . The same group reported very recently that lipoglycans of the Mtb cell wall stimulate the activation of NF-kB via TLR-2 and that the subsequent upregulation of cellular FLIP leads to inhibition of FasL-mediated apoptosis [24] . Furthermore , it was suggested that Mtb stimulated the secretion of soluble TNF-R2 , which led to the reduction of bioactive TNF-α in the medium and therefore less stimulation of the TNF-R1 [25] . Altogether , it seems that virulent Mtb is able to inhibit induction of host cell apoptosis via multiple pathways , and probably encodes mechanisms to interfere with both intrinsic and extrinsic pathways for initiation of programmed cell death . The inhibition of macrophage apoptosis by Mtb is believed to provide a number of advantages to the bacterium in its struggle to resist the host immune response . These include preservation of a favorable host cell environment during growth and persistence [26 , 27] , evasion of apoptosis-linked bactericidal effects [28 , 29] , and avoidance of efficient cytotoxic T cell priming via the detour pathway of antigen cross-presentation [15 , 30–32] . This last point is of potential importance to the improvement of tuberculosis vaccines , because attenuated mycobacterial strains that induce higher levels of host cell apoptosis would be expected to stimulate more robust cellular immunity , as suggested by a recent study using recombinant M . bovis Bacille Calmette-Guérin ( BCG ) expressing listeriolysin [33 , 34] . Therefore , the identification of mycobacterial genes required for prevention of apoptosis could lead to specific strategies for designing more efficacious forms of BCG or other attenuated mycobacterial vaccine strains . In order to clarify the role of mycobacteria in host cell apoptosis and to address its importance for bacterial virulence , we sought to identify anti-apoptosis genes via a gain-of-function genetic screen . Using this approach , we successfully identified two independent genomic regions of virulent Mtb ( strain H37Rv ) that mediate the inhibition of host cell apoptosis by the facultative pathogen M . kansasii . The analysis of a defined set of bacterial mutants of M . kansasii in immunocompromised ( SCID ) mice demonstrated a causal relationship between inhibition of apoptosis and virulence . These findings were confirmed via a loss-of-function strategy using the newly identified anti-apoptosis gene , nuoG , and demonstrated attenuation of Mtb nuoG mutants in immunocompromised and immunocompetent mice . Altogether our findings allowed , to our knowledge for the first time , the demonstration of a causal relationship between inhibition of host cell apoptosis and virulence of mycobacteria .
To identify genes in Mtb responsible for anti-apoptotic effects , we established a gain-of-function genetic screen using the nonpathogenic M . smegmatis mc2155 strain , which is a fast-growing mycobacterium that is extremely efficient for transformation [35] . The human cell line THP-1 was chosen for use as the host cells for this screen because published work indicates that these cells provide an accurate model for the apoptotic response of Mtb-infected primary human alveolar macrophages [23] . Our initial studies established that M . smegmatis infection of THP-1 cells induced strong apoptosis after 1 . 5 d of infection when compared to BCG-infected macrophages , as assessed by disruption of the cell monolayer and by staining for DNA strand breakage using the TUNEL assay ( Figure 1A ) . The failure of BCG to induce apoptosis seems to contradict the report by Kean et al . [17] in which seven different species of mycobacteria were compared for their capacity to induce apoptosis in primary human alveolar macrophages , and BCG was among the apoptosis-inducing mycobacterial species . Nevertheless , in that study , no fast-growing species were included , and all apoptosis assays were performed after 5–7 d of infection . In an independent study , we demonstrated that M . smegmatis and other nonpathogenic mycobacteria like M . fortuitum have a very strong capacity to induce apoptosis with rapid kinetics , even when compared to facultative pathogenic mycobacteria like M . kansasii and BCG ( A . Bohsali and V . Briken , unpublished data ) . Therefore , the induction of apoptosis by M . smegmatis had to be analyzed very early after infection ( 16–36 h; Figure 1A and 1B ) , and at this time point , little or no induction of macrophage apoptosis by BCG could be observed . The gain-of-function screen was performed using a library of 312 M . smegmatis clones containing Mtb genomic DNA fragments on an episomal cosmid . Two clones ( designated J21 and M24 ) containing cosmids with separate nonoverlapping Mtb genomic DNA inserts gave significantly reduced levels of apoptosis and were selected for detailed study . To confirm that the observed effects on apoptosis were due to the cosmids contained in these clones , the episomal cosmid DNA was extracted and re-transfected into M . smegmatis . In both cases , the re-transformed clones had the same phenotype as the original clones , showing an approximately 50% reduction of apoptosis of infected THP-1 cells ( Figure 1B ) . It is important to note that our screening was designed to emphasize specificity rather than sensitivity , as it focused on the clones showing the strongest and most reproducible suppression of apoptosis . Thus , it is very likely that other regions with anti-apoptosis capacity may remain to be identified . Although the effects of cosmids M24 and J21 on reducing the apoptosis induced by M . smegmatis were highly reproducible , the magnitude of this effect was relatively modest . Most likely , this reflected the very strong capacity of M . smegmatis to induce apoptosis when compared to other mycobacteria ( Figure 1 and unpublished data ) , and we assessed this by testing the effects of the Mtb cosmids on apoptosis induction by another mycobacterial species , M . kansasii . This opportunistic pathogen is known to be a strong inducer of apoptosis , but it shows slower host cell killing than M . smegmatis , with significant levels of apoptosis being observed only after 5–7 d of infection [17] . We transformed the cosmids into M . kansasii to generate clones Mkan-J21 and Mkan-M24 , and also generated an M . kansasii control strain using the empty cosmid vector pYUB415 ( Mkan-CO ) . Based on FACS analysis of TUNEL staining , wild-type M . kansasii and Mkan-CO induced comparable high levels of apoptosis at 5 d post infection ( 95% and 86% apoptosis , respectively ) , whereas Mkan-J21 and Mkan-M24 showed markedly reduced levels ( 16% and 19% apoptosis , respectively ) ( Figure 1C ) . The transfection of the two cosmid clones did not affect the in vitro growth of either M . smegmatis or M . kansasii ( unpublished data ) . Apoptosis of infected macrophages has been reported to directly kill ingested bacteria [29] , and killing of bacteria within apoptotic bodies is also facilitated as a result of enhanced phagocytosis by activated bystander macrophages [28] . Therefore , we hypothesized that inhibition of apoptosis is important for mycobacterial evasion of the host's innate immune response . This was tested by infecting groups of SCID mice ( BALB/c background ) with the M . kansasii cosmid transformants or with Mtb H37Rv . As expected , H37Rv was highly virulent in these mice that lack adaptive immunity ( median survival 15 d ) , whereas Mkan-CO showed only modest virulence even in these immunodeficient animals ( median survival >200 d ) ( Figure 2A ) . Remarkably , Mkan-J21 and Mkan-M24 showed significantly increased virulence in SCID mice , with median survival times of 44 d or 60 d , respectively ( p = 0 . 0002 compared to Mkan-CO for both survival times by log-rank test ) . Tissue bacterial burdens in SCID mice infected with the various mycobacteria were consistent with the survival data , based on colony-forming units ( CFUs ) in the lung , liver , and spleen at various time points after infection ( Figure 2B ) . Histopathology of the lungs revealed that after 35 d , airways in the lungs of Mkan-J21–infected mice were almost completely consolidated , and in Mkan-M24–infected mice the lungs showed significant infiltration of inflammatory cells . In contrast , lungs of mice infected with Mkan-CO had normal morphology ( Figure 2C ) . In order to correlate the virulence of the different strains with their capacity to inhibit apoptosis , lung sections obtained 14 d after infection were stained for apoptotic cells using a TUNEL-based assay and analyzed by microscopy as described in the Materials and Methods section . This revealed minimal levels of apoptosis in situ for lung tissue infected with H37Rv ( 2% ) , Mkan-J21 ( 5% ) , and Mkan-M24 ( 7% ) compared to significantly higher levels in Mkan-CO–infected ( 27% ) lungs ( Figures 2D and S1 ) . Overall , these results confirmed the anti-apoptotic activity of genes contained in cosmids J21 and M24 , and strongly supported the importance of apoptosis inhibition for the virulence of mycobacteria . Cosmids J21 and M24 both contained about 30 mycobacterial genes , and therefore it could not be completely excluded that the effect on bacterial virulence was in part or totally caused by another gene or genes linked to those responsible for the anti-apoptosis effects . As Mkan-J21 showed the strongest enhancement of virulence in SCID mice , we selected cosmid J21 for additional studies to determine the precise gene or genes required for inhibition of apoptosis by M . tuberculosis . Sequencing of J21 cosmid DNA ( unpublished data ) showed that its insert corresponded to bp 3511794 through bp 3545572 of the Mtb genome , according to the standard annotation for strain H37Rv [36] . This interval contains the intact open reading frames of 31 annotated genes , including a large operon that encodes the 14 subunits of the Mtb type I NADH dehydrogenase complex ( NDH-1 ) , along with other genes that encode a variety of different known or predicted functions ( Figure S2 ) . To identify which gene or genes in this region were important for the anti-apoptosis activity , a series of deletion mutants spanning different regions within the genomic interval corresponding to J21 was created in Mtb H37Rv using specialized transduction [37] ( Figures S2 and S3 ) . These mutants were analyzed by using SCID mice to test for reduction in bacterial virulence ( Figure S4A ) , and by using THP-1 cells to test for loss of apoptosis inhibition ( Figure S4B ) . Although several of the deletions showed modest effects , only one mutant , Mtb ΔRv3151 , which corresponded to the deletion of the nuoG subunit of NDH-1 , gave both statistically significant extension of survival in SCID mice and enhanced apoptosis in THP-1 cells relative to parental Mtb . The gain-of-function experiments described in Figures 1 and 2 using the episomal cosmids relied on the proper function of the endogenous gene promoters in transfected M . smegmatis or M . kansasii . To confirm that nuoG was actually transcribed on the J21 cosmid in M . smegmatis , reverse transcription–PCR with primers specific for Mtb nuoG was performed and clearly demonstrated that the nuoG gene was transcribed ( Figure S4C ) . These experiments implicated nuoG , and potentially the complete functional NDH-1 complex , in mediating most or all of the anti-apoptotic properties of cosmid clone J21 . To confirm that nuoG deletion was responsible for the pro-apoptotic phenotype and to exclude significant polar effects of the deletion on other nearby genes , the mutant was complemented with a plasmid carrying a copy of nuoG behind a constitutively active promoter that integrates into the Mtb chromosome at the unique attB site [38] . This gave full complementation for the in vitro apoptosis assays ( Figure 3 ) , although there was a residual increase in apoptosis induction observed in vivo ( Figure 4B and 4F ) that may have been due to some minor polar effects on the transcription of other members of the nuo-operon . The mutant ( MtbΔ ) , complemented mutant ( MtbΔC ) , and wild-type Mtb were analyzed for the capacity to grow in vitro , which demonstrated that the nuoG deletion had no effect on aerobic growth rate ( Figure 3A ) , confirming a previous report that NDH-1 is not essential for mycobacterial growth in culture [39] . This is most likely due to the fact that Mtb has two additional NADH dehydrogenases , which are the non-proton pumping type II NADH dehydrogenases encoded by the ndh and ndhA genes [40] . To confirm the pro-apoptotic effect of nuoG deletion , we compared levels of apoptotic cell death in cultures of differentiated THP-1 cells and extended the analysis to include detection of apoptosis in bone marrow–derived macrophages ( BMDMs ) from BALB/c mice following infection . The Mtb nuoG mutant induced significantly more apoptosis than the complemented strain or the wild-type Mtb in human and mouse macrophages ( Figure 3B and 3C ) . This phenotype of the nuoG mutant was also confirmed in BMDMs from C57BL/6 mice ( unpublished data ) . The difference in apoptosis induction was not due to a reduced phagocytosis of the nuoG mutant , which was demonstrated by comparing rates of infection via acid-fast staining ( Figure 3D ) and CFU determination ( unpublished data ) . These results show that nuoG was necessary for Mtb to inhibit apoptosis of primary murine macrophages or the human macrophage-like THP-1 cells . The nuoG Mtb mutant still induced less apoptosis ( Figure 3B ) when compared to M . kansasii–infected cells ( Figure 1C ) , which probably reflects the fact that virulent Mtb expresses multiple anti-apoptosis genes . Therefore , the deletion of only one gene does not completely abolish the capacity of the bacterium to inhibit apoptosis . Since nuoG is part of a multi-subunit NDH-1 complex that is present in the cosmid J21 , it was of interest to determine if nuoG alone could confer the gain-of-function in M . kansasii or if , in contrast , expression of the whole nuo-operon was needed . Therefore , Mtb-nuoG was constitutively expressed in M . kansasii and the capacity to induce host cell apoptosis was analyzed ( Figure 3E ) . This demonstrated that expression of Mtb-nuoG alone very significantly reduced apoptosis induction by M . kansasii when compared to that of wild-type and empty vector-transfected bacteria ( Figure 3E ) . Most of the sequenced genomes of mycobacteria contain a nuoG gene within a nuo-operon containing 14 genes that code for the NDH-1 . The one exception to date is M . leprae , in which the whole operon is deleted except for a nuoN pseudogene [41] . nuoG of M . kansasii was cloned using PCR and sequenced in order to allow protein sequence comparison of all the nuoGs in the mycobacterial strains used in our study . Comparison of the nuoG protein sequences among virulent mycobacteria revealed a high degree of homology ( 99% identity , Figure S9 ) . Interestingly , nuoG of BCG is also highly homologous ( 99% ) to nuoG of virulent mycobacterial species , which is consistent with our unpublished data demonstrating that the deletion of nuoG in BCG also increases the potential of the bacteria to induce apoptosis . These findings suggest that the vaccine strain BCG retained this virulence mechanism from its parental M . bovis strain , although overall it may still induce more apoptosis than fully virulent mycobacteria . In contrast , nuoG of M . smegmatis is only 70% identical ( Figure S9 ) and nuoG of M . kansasii is only 34% identical to nuoG sequences of virulent mycobacterial species . The latter nuoG protein is truncated due to a stop codon introduced at codon 295 , and thus nuoG of M . kansasii is missing about 512 of the C-terminal amino acids . Interestingly , the M . kansasii nuoG is quite homologous up to amino acid 288 ( 94% ) . In conclusion , it seems likely that both nuoG of M . smegmatis and M . kansasii have lost their apoptosis-inhibiting function within the NDH-1 complex , and therefore the overexpression of Mtb nuoG is able to restore the capacity of the bacteria to inhibit host cell apoptosis . It will be of great interest to explore this hypothesis further by examining the expression of the various nuoG proteins in the nuoG deletion mutant of Mtb and analyzing the apoptosis induction of these complemented bacteria . How does nuoG mediate apoptosis inhibition ? For nuoG to have a direct effect on host cell apoptosis pathways , one would assume that it needs to be secreted in order to interact with host cell proteins or lipids . However , if the structure of the NDH-1 complex of Mtb is similar to that of other bacterial NDH-1 complexes , nuoG will be located in the cytosol of the bacterium [42] . To determine whether Mtb nuoG is secreted experimentally , we created a phoA-nuoG fusion protein . phoA can convert a colorless substrate into a blue product , but only if it is secreted by the bacterium [43] . This assay failed to detect secretion of nuoG ( Figure S7 ) , which is consistent with the absence of a signal peptide and the predicted cytosolic localization of this component of NDH-1 . Altogether , it thus seems unlikely that nuoG is secreted , and a direct effect of nuoG onto the host cell can also be judged to be unlikely . The disruption of the NDH-1 system in the nuoG mutant might have a very profound impact on the metabolism and proteome of the mycobacterium , which might result in an indirect effect on host cell apoptosis induction . Nevertheless , the absence of an in vitro growth defect in the nuoG mutant ( Figure 3A ) would argue against a profound effect of the deletion on bacterial metabolism . In order to address the effect of the nuoG mutation on the proteome of the mycobacteria , the proteins of wild-type and mutant Mtb were separated via 2-D gel electrophoresis . This revealed that the nuoG mutation did not induce a major change in the proteome ( Figure S8 ) . Thus , we found that the deletion of nuoG did not have a major impact on the general bacterial metabolism or its proteome . Instead , we propose that nuoG exerts its anti-apoptotic and virulence-promoting function via the enzymatic activity of the NDH-1 complex in a more specific way . Regardless of the remaining questions about the potential mechanism of the nuoG/NDH-1–mediated apoptosis inhibition , our identification of an apoptosis-inducing mutant of Mtb allowed us to analyze the importance of apoptosis inhibition for bacterial virulence . First , the importance of host cell apoptosis inhibition in innate immune defense was analyzed by infecting immunodeficient SCID mice . The median survival times for mice infected with Mtb wild-type or the complemented strain were not significantly different ( 14 d and 16 . 5 d , respectively ) . In contrast , mice infected with ΔnuoG Mtb survived twice as long as those infected with wild-type bacteria ( median survival of 27 . 5 d , p < 0 . 0001 , log-rank test; Figure 4A ) , even though all mice received similar initial bacterial doses as confirmed by CFU counts at day 1 after infection ( Figure S10 ) . Consistently , the amount of apoptosis induced in lung sections of these mice was significantly increased from less than 1% with wild-type bacteria to about 13% in mutant bacteria ( p < 0 . 05 ) ( Figures 4B and S5 ) . The complemented bacteria still showed increased apoptosis ( 6% ) , but this was significantly reduced compared to the nuoG mutant ( p < 0 . 05 ) . Therefore , the deletion of nuoG significantly increased the induction of apoptosis in the lungs of SCID mice . These results corroborated the findings presented in Figure 2 by using a loss-of-function approach , and together both sets of experiments point towards an important role of infection-induced apoptosis in the innate immune response . This statement is supported by the recent results linking the capacity of host cell macrophages to undergo apoptosis upon mycobacterial infection to the susceptibility of different mouse strains [20] . The importance of nuoG-mediated apoptosis inhibition for bacterial virulence in immunocompetent mice was analyzed using BALB/c mice ( Figure 4C–4F ) . Again , the mutant significantly ( p < 0 . 004 ) delayed the death of infected mice when compared to wild-type ( median survival 175 d ) and complemented bacteria ( median survival 193 d , p = 0 . 16 compared to wild-type ) ( Figure 4C ) . Measurement of bacterial CFUs showed that the growth of the mutant in the lungs of BALB/c mice was significantly reduced by approximately 0 . 8 log at week 20 , after similar initial growth during the first 3 wk of infection ( Figure 4D ) . In contrast , the bacterial loads in spleen and liver were not significantly different at 3 , 10 , or 20 wk post infection ( unpublished data ) . Comparison of the histopathology of lung sections of infected BALB/c mice demonstrated an obvious reduction in granulomatous inflammation at week 20 in animals infected with the ΔnuoG mutant ( Figure 4E ) . Although lung histopathology appeared similar in wild-type and ΔnuoG mutant–infected BALB/c mice at 3 wk ( Figure 4E ) , staining of lung sections at this time point for TUNEL reactivity revealed a significant increase of greater than 10-fold in apoptotic cells in mice infected with the ΔnuoG mutant compared to mice that were uninfected , or infected with wild-type or complemented bacteria ( Figures 4F and S6 ) . Taken together , our results demonstrate that nuoG is an anti-apoptosis gene of Mtb that is important for bacterial virulence in both immunocompromised and immunocompetent mice and thus strongly support the general hypothesis that the inhibition of host cell apoptosis is important for virulence of mycobacteria . The challenge ahead is to determine the molecular mechanism by which a bacterial NADH dehydrogenase can manipulate host cell apoptosis induction . It is intriguing to speculate that perhaps the NDH-1 complex of virulent bacteria has taken on a separate function in modifying apoptotic responses of infected macrophages from its original purpose of energy generation , which is now mainly performed by the type II dehydrogenases ndh and ndhA , which are both essential genes . One of the unique features of NDH-1 , as opposed to the NDH and NDHA dehydrogenases of M . tuberculosis , is its capacity to pump protons across the bacterial membrane . We therefore hypothesize that these protons , in conjunction with the secreted bacterial superoxide dismutases ( SodA and SodC ) , could serve to neutralize the superoxide anions generated within the phagosome by an activated NOX2 complex to generate hydrogen peroxide , which is then further catabolized to water and oxygen by bacterial catalase ( KatG ) . Superoxide anions are a known trigger for apoptosis in a variety of biological systems , so the involvement of NDH-1 and nuoG in their elimination may interrupt a critical signal that initiates the host cell apoptosis response . Our hypothesis would predict that other NDH-1 subunits involved in proton translocation , such as nuoL and nuoM [42] , will have the same apoptosis phenotype as the nuoG mutant , a prediction that has not yet been tested experimentally . This proposed general mechanism for inhibition of apoptosis is further supported by other studies implicating SodA as an anti-apoptotic factor in Mtb ( [44 , 45] ) . The discovery of nuoG and NDH-1 as anti-apoptosis factors encoded by specific Mtb genes suggests new strategies for improving currently used and novel tuberculosis vaccines , and could also provide targets for development of antimicrobial drugs for treatment of persistent disease .
M . smegmatis ( mc2155 ) has been previously described [35] , and M . kansasii strain Hauduroy ( ATCC 12478 ) and Mtb strain H37Rv ( ATCC 25618 ) were obtained from the American Type Culture Collection ( http://www . atcc . org/ ) . M . bovis BCG Pasteur strain was obtained from the Trudeau Culture Collection ( Saranac Lake , New York , United States ) . GFP-expressing BCG and M . smegmatis were generated by subcloning the enhanced GFP gene ( Clontech , http://www . clontech . com/ ) into the mycobacterial episomal expression vector pMV261 . The resulting plasmid ( pYU921 ) was transfected into competent cells by electroporation as previously described [35] . M . smegmatis was cultured in LB broth with 0 . 5% glycerol , 0 . 5% dextrose , and 0 . 05% TWEEN-80 . Mtb H37Rv and M . kansasii were grown in 7H9 broth with 0 . 5% glycerol , 0 . 5% dextrose , 0 . 05% TWEEN-80 , and 10% OADC enrichment ( DIFCO , http://www . bd . com/ds/ ) . For selective media , 50 μg/ml hygromycin or 40 μg/ml kanamycin were added . Human myelomonocytic cell line THP-1 ( ATCC TIB-202 ) was cultured and differentiated using phorbol myristate acetate ( PMA ) ( Sigma , http://www . sigmaaldrich . com/ ) as described [46] . Bone marrow macrophages were derived from the femur and tibia of BALB/c mice as described [46] . Bacteria were grown to an OD600 ranging from 0 . 5 to 0 . 8 , sonicated twice for 20 s using a cup horn sonicator , and allowed to settle for 10 min . The infection was carried out at a multiplicity of infection ( MOI ) of 10:1 ( 10 bacilli to 1 cell ) for 4 h in triplicate wells , after which extracellular bacteria were removed by four washes with phosphate buffered saline ( PBS ) . The cells were incubated in DMEM ( Invitrogen , http://www . invitrogen . com/ ) with 20% human serum ( Sigma ) and 100 μg/ml gentamicin ( Invitrogen ) , and an apoptosis assay was performed after the indicated periods of culture . The TUNEL assay was performed to reveal apoptosis-induced DNA fragmentation in either tissue culture cells or lung sections of infected mice using the In Situ Cell Death Detection Kit , Fluorescein ( for cultured cells ) or In Situ Cell Death Detection Kit , POD ( for lung sections ) ( Roche Applied Science , http://www . roche-applied-science . com/ ) . The assay was carried out as described by the manufacturer and the percentage of stained cells was analyzed using flow cytometry for cultured cells or quantification via light microscopy for the animal tissue sections . The strategy for generation of the Mtb genomic library in cosmid vector pYUB415 has been previously described [47] . Briefly , Mtb ( strain Erdman ) genomic DNA was purified and partially digested with Sau3A . DNA fragments of about 40 kbp were selected by agarose gel purification and ligated into arms of cosmid vector pYUB415 digested with BamH1 as previously described [47] . DNA was packaged in vitro with Gigapack XL ( Stratagene , http://www . stratagene . com/ ) and Escherichia coli were transduced and selected on LB plates containing 100 μg/ml ampicillin . Over 105 independent clones were pooled , and DNA for transformation was obtained using standard alkaline lysis method . Transformations were performed by electroporation of competent mycobacteria as described [35] . For the initial screen , M . smegmatis was transformed with the genomic DNA cosmid library described above , and 312 cosmid clones were picked and grown in liquid medium containing 50 μg/ml hygromycin . Assuming random distribution of Mtb sequences among the cosmid transformants and an average insert size of about 40 kbp , the 312 cosmid clones represented 3-fold coverage of the entire Mtb genome . After three successive screens using bright field microscopy or flow cytometry to assess levels of cell death , 12 clones were selected for quantitative assessment using TUNEL staining followed by flow cytometry . This identified three clones of greatest interest ( M24 , J21 , and I16 ) , and their cosmid DNA was purified and screened by restriction digest ( not shown ) . This revealed that the inserts of M24 and I16 were identical , but different from the insert of J21 . For cosmid J21 , the 5′ and 3′ ends of the insert DNA were sequenced and aligned with the published genomic Mtb DNA sequence ( Figure S2 ) , and subsequently the whole insert was sequenced to confirm that is corresponded to the sequence published data . Specific genes of Mtb were disrupted using specialized transduction as described [37] . To create the nuoG::hyg-null allele , the hygromycin resistance cassette was introduced between the first 4 bp of the nuoG 5′ end and the last 163 bp of the 3′ end of the open reading frame . The successful deletion of the gene was demonstrated by Southern blotting as described previously ( Figure S3 ) . For complementation of the ΔnuoG mutation , nuoG was amplified by PCR and cloned behind the hsp60 constitutive promoter into the plasmid pMV361 , which allows integration of a single copy into the genome of Mtb [38] . BALB/c or SCID/Ncr ( BALB/c background ) mice ( 4- to 6-wk-old females ) were infected intravenously through the lateral tail vein with 1 × 106 bacteria . For survival studies , groups of ten mice were infected , and after 24 h , three mice per group were sacrificed to determine the bacterial load in the organs . In order to follow the bacterial growth , an additional three mice per time point were infected . The organs ( lung , spleen , liver ) were homogenized separately in PBS/0 . 05% TWEEN-80 , and colonies were enumerated on 7H10 plates grown at 37 °C for 3–4 wk . For histopathology , tissues were fixed in 10% buffered formalin and embedded in paraffin; 4-μm sections were stained with hematoxylin and eosin . TUNEL staining was performed on the paraffin-embedded tissue sections using the In Situ Cell Death Detection Kit , POD ( Roche Applied Science ) per the manufacturer's protocol . Quantification was performed on coded specimens by a blinded observer by counting the number of apoptotic nuclei per ∼200 total nuclei in eight separate areas of two lung sections for each of the three mice per group . All animals were maintained in accordance with protocols approved by the Albert Einstein College of Medicine and University of Maryland Institutional Animal Care and Use Committees .
The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank/index . html ) accession numbers for the sequences of the nuoG proteins from the following mycobacteria are M . bovis AF2122/97 ( CAD95267 ) , M . bovis BCG-Pasteur ( YP979258 ) , M . kansasii Hauduroy ( EF607211 ) , M . smegmatis mc2155 ( YP886418 ) , M . tuberculosis CDC1551 ( AAK47578 ) , and M . tuberculosis H37Rv ( CAB06288 ) . | The infection-induced suicide of host cells following invasion by intracellular pathogens is an ancient defense mechanism observed in multicellular organisms of both the animal and plant kingdoms . It is therefore not surprising that persistent pathogens of viral , bacterial , and protozoal origin have evolved to inhibit the induction of host cell death . M . tuberculosis , the etiological agent of tuberculosis , has latently infected about one third of the world's population and can persist for decades in the lungs of infected , asymptomatic individuals . In the present study we have identified nuoG of M . tuberculosis , which encodes a subunit of the type I NADH dehydrogenase complex , as a critical bacterial gene for inhibition of host cell death . A mutant of M . tuberculosis in which nuoG was deleted triggered a marked increase in apoptosis by infected macrophages , and subsequent analysis of this mutant in the mouse tuberculosis model provided direct evidence for a causal link between the capacity to inhibit apoptosis and bacterial virulence . The discovery of anti-apoptosis genes in M . tuberculosis could provide a powerful approach to the generation of better attenuated vaccine strains , and may also identify a new group of drug targets for improved chemotherapy . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"infectious",
"diseases",
"none",
"in",
"vitro",
"immunology",
"microbiology",
"mus",
"(mouse)",
"eubacteria"
] | 2007 | Mycobacterium tuberculosis nuoG Is a Virulence Gene That Inhibits Apoptosis of Infected Host Cells |
Intelligent organisms face a variety of tasks requiring the acquisition of expertise within a specific domain , including the ability to discriminate between a large number of similar patterns . From an energy-efficiency perspective , effective discrimination requires a prudent allocation of neural resources with more frequent patterns and their variants being represented with greater precision . In this work , we demonstrate a biologically plausible means of constructing a single-layer neural network that adaptively ( i . e . , without supervision ) meets this criterion . Specifically , the adaptive algorithm includes synaptogenesis , synaptic shedding , and bi-directional synaptic weight modification to produce a network with outputs ( i . e . neural codes ) that represent input patterns proportional to the frequency of related patterns . In addition to pattern frequency , the correlational structure of the input environment also affects allocation of neural resources . The combined synaptic modification mechanisms provide an explanation of neuron allocation in the case of self-taught experts .
Adaptive synaptogenesis [1–4] is designed to allocate neural resources in a thrifty manner or in a manner to regulate function . The three resources of concern are number of synapses , number of neurons , and firing-rate of the neurons . Inspired by the Bienenstock-Cooper-Munro ( BCM ) algorithm [5] and its forcing of a neuron to a predefined activity level , adaptive synaptogenesis achieves a similar goal that not only guarantees the average activity of a postsynaptic neuron but does so in a way that rations synapses . Previously , adaptive synaptogenesis was used as a mechanism to produce compressive coding with small information losses [6–10] . It also successfully models developmental studies of ocular dominance [11–12] . Both results are achieved by postsynaptic neurons discovering implicit correlational structures within the input data space . Through the random acquisition and forced shedding of synapses , associated inputs find their way to the same neuron , and uncorrelated or anti-correlated inputs are ignored . As thus conceived , adaptive synaptogenesis consists of ( 1 ) a random Bernoulli process that selects a new excitatory connection between nearby axon i and postsynaptic neuron j; ( 2 ) once formed , associative synaptic modification controls the strength of each existing synapse , and this control includes the possibility of potentiation , depression , or no change of a synaptic weight; but with enough long-term depression , ( 3 ) shedding of a synapse occurs when the weight is appropriately weak ( near zero for a sufficiently long time ) ( Fig 1 ) . Critically , the possibility of forming a new synapse on neuron j is determined by j’s long-term average firing-rate . Instead of compressive coding , the context for studying adaptive synaptogenesis here is self-taught discrimination . The motivating idea is that if one studies a particular field and its subject matter over a long enough period of time ( perhaps the oft quoted ten thousand hours [13] ) and if one studies over a wide enough variety of representative examples , the allocation of neurons in the cerebral cortex is enhanced for this particular concentrated field of study . After a detailed description of the neural algorithm and the input data structures , we establish a mathematical theory that quantifies relationships ( e . g . synaptic weights ) required for stability ( or lack thereof ) of the neurons formed by this algorithm . Computational simulations follow these theoretical developments . The simulations demonstrate the development of stable neuron configurations without turning-off the algorithm . Moreover , these simulations also reveal the effect of input statistics—frequency of input patterns and the input correlational structure—on neuron allocation . As shown , the form of adaptive synaptogenesis used here produces neuron allocations that are appropriately biased by the statistics of the input environment ( more experience produces more neurons devoted to the experience ) . Also revealed is an important effect of the input world’s statistical structure that can help or hinder this proportional neuron allocation .
Here we study an adaptively constructed , feedforward network of McCulloch-Pitts neurons . The inputs are vectors with binary elements , xi ( k ) ϵ{0 , 1} , and the outputs are vectors with binary elements , zj ( k ) . For the jth neuron , postsynaptic excitation is linear , yj ( t ) = Σixi ( t ) ·cij ( t ) ·wij ( t ) with connection indicator cij ( k ) ϵ{0 , 1} , with all weights wij positive , and output zj ( k ) : = {1 if yj ( k ) ≥θ , and 0 otherwise} . Threshold θ is 3 . 0 for dataset A and 0 . 8 for datasets B . The “sensory” input dimensions are 80 ( dataset A ) or 390 ( datasets B ) as described below . The number of postsynaptic neurons simulated is 2000 per dataset . Because there is no interaction between the outputs of these neurons ( i . e . there is no feedback or lateral inhibition ) and because there is no avidity rule [7] , each neuron develops its connections independent of all other neurons . Each neuron is initialized with one connection from a randomly chosen input line with weight 0 . 2 . There are three distinct aspects of synaptic modification: synaptogenesis , associative synaptic modification , and synaptic shedding . A connection from input neuron i to output neuron j is indicated as cij ( t ) ϵ{0 , 1} . The strength ( weight ) of this connection is wij ( t ) ϵ ( 0 . 01 , 1 ) . Inputs are binary , i . e . , xi ( t ) ϵ{0 , 1} . Excitation of a neuron on a timestep is linear , yj ( t ) = Σicij ( t ) ·wij ( t ) ·xi ( t ) . There is no inhibition . A neuron , whose excitation reaches threshold , fires by the rule zj ( t ) = {1 if yj ( t ) ≥θ , and 0 otherwise} . A weight is updated according to [15]; in particular , Δwij ( t ) = ε·cij ( t ) · ( xi ( t ) – E[xi] – wij ( t ) ) ·yj ( t ) . On each timestep , a neuron’s moving-average of firing-rate , z¯j , is updated as z¯j ( t ) =z¯j ( t−1 ) ⋅ ( α ) +zj ( t−1 ) ⋅ ( 1−α ) . Synaptogenesis is controlled by z¯j and a random variable: uij ϵ{0 , 1} where prob ( uij = 1 ) = γ . A connection/weight is shed whenever it falls below 0 . 01; that is , if ( wij ( τ ) <0 . 01 ) , then cij ( τ+1 ) =0 & wij ( τ+1 ) =∅ . There are two sets of parameterizations . There is one parameterization for dataset A and one parameterization for datasets B1 , B2 , and B3 . However , many parameterizations were examined for each dataset , and in fact there are ranges of parameter settings for which the generic results presented below are valid . In this case ‘valid’ means a parameter set that produces stable connections and postsynaptic neurons that exceed the desired minimum firing-rate . For the results presented here , the parameterizations produce a relatively large number of synapses per neuron compared to other valid settings . The parameterizations are listed in Table 1 . Neuron parameterizations that change between dataset A and datasets B1 , B2 , and B3 are threshold to fire ( 3 . 0 versus 0 . 8 ) and minimum desired firing-rate ( 9% versus 10% ) . As an explicit part of the model , there are three time-scales: i ) the shortest is the neuron update ( fire or not , given an input ) ; ii ) the next in duration is synaptic modification of existing synapses , which occurs every timestep; and iii ) the slowest time-scale , which occurs after each training block , synaptogensis and shedding . In one cycle , all input vectors are presented to the network . A training block of inputs equals 10 cycles ( e . g . if there are 50 input states , a block occurs after the network is presented with 500 inputs ) . The length of the simulations are shown in Table 2 . For each postsynaptic neuron the input blocks are repeated until no synapses are gained or lost for 200 blocks . At this time , a neuron’s synapses are assumed stable . At this point , as shown in the results , the synaptic weights have achieved their predicted values . S1 Code contains the Matlab program used for simulations . There are 2000 neurons simulated per dataset; this large number serves the purpose of producing accurate statistics . However , because the synaptic modification algorithms used here yield feed-forward networks with excellent data compression and little information loss , certain analyses only make sense when the number of neurons are much fewer than 2000; in particular we limit the number of randomly sampled neurons to 1 through 50 out of the 2000 . Neuron construction is driven by repeated presentation of patterns from a predetermined dataset . Four different input environments are studied ( Table 3 ) . The first dataset ( dataset A , see Fig 2A and S1 Dataset ) has 80 input lines . There are five orthogonal prototypes that define the corresponding categories: the five prototypes correspond to a higher probability of firing within one of five distinct sets of input lines 1–16 , 17–32 , 33–48 , 49–64 , or 65–80 . The five exemplars are generated from these prototypes , and presented with relative frequency 0 . 1 , 0 . 15 , 0 . 2 , 0 . 25 , and 0 . 3 for a total of 100 input patterns . In the case of dataset A , each prototype is randomly perturbed such that the total number of active input-lines for each generated pattern remains constant . Specifically , two randomly selected active input-lines of the prototype are inactivated , and two randomly selected inactive input-lines of the prototype are activated . Fig 2A illustrates the binary input vectors of dataset A . Note the small amount of overlap between the input patterns only occurs due to noise . Dataset B1 is much more complex in terms of relationships between input vectors . This set consists of nine categories each with its own prototype . Each of the nine individual categories has the same relative frequency ( 11 . 1% ) . The nine categories can be partitioned evenly into three super-categories: I , II , and III . Fig 2B visualizes the input set . These inputs are coded as 390-dimensional binary vectors . Between super-categories , the input patterns are orthogonal . Within a super-category , the prototypes and the patterns they generate overlap; the degree of overlap varies as a function of the super-category . Within the three super-categories overlap increases from 5 to 10 , and then 15 input lines for super-categories I , II , and III , respectively . Each super-category has some input lines that are activated by only a single category; other input lines of this super-category are shared by two of the three categories; and the remaining input lines of the super-category are shared by all three categories . Super-category I has the least overlap: category A has 45 potentially active input lines that belong only to category A , 5 that belong to A and B , 5 that belong to A and C , and 5 that belong to A , B , and C for a total of 60 potentially active input lines per category . Super-category II has more overlap between its three categories ( D , E , and F ) . There are 30 potentially active input lines that belong only to category D , 10 that belong to D and E , 10 that belong to D and F , and 10 that belong to D , E and F for a total of 60 potentially active input lines . Super-category III has the most overlap . There are 15 potentially active input lines that belong only to category G , 15 that belong to G and H , 15 that belong to G and J , and 15 that belong to G , H , and J for a total of 60 potentially active input lines . Thus , each category has a total of 60 potentially active input lines . Exactly 20 of the 60 potentially active input lines from each category are pseudo-randomly chosen to be active for each pattern . None of the super-category’s designated inactive input lines are turned into active input lines . Datasets B1 , B2 , and B3 are carefully constructed to illustrate the effects of category overlap ( Fig 3 and S1 Dataset ) versus category probability on neuron allocation . Datasets B2 and B3 are constructed in a similar way as dataset B1 , but with different relative frequencies for each category . Dataset B2 has relative frequencies: 0 . 13 , 0 . 13 , 0 . 13 , 0 . 11 , 0 . 11 , 0 . 11 , 0 . 098 , 0 . 093 , and 0 . 093 . Dataset B3 has relative frequencies: 0 . 18 , 0 . 17 , 0 . 15 , 0 . 12 , 0 . 11 , 0 . 087 , 0 . 063 , 0 . 058 , and 0 . 053 .
The most important idea of this section is that there is a mathematical derivation that characterizes the stable connectivities for a feedforward neuron whose connections are governed by adaptive synaptogenesis . This theory’s convergence results provide a means for identifying stable configurations when simulations are performed . Going in the other direction , and of secondary importance is establishing the relevance of the theory via simulations because the theory requires multiple infinities of sampling: thus , simulations must be used to establish the existence of parameterizations capable of achieving the predicted , stable connectivities . The stable weight values on a neuron with a stable connectivity are a function of the subspace covariance matrix that arises from the set of input lines received by this neuron . For example , one of our input environments is a random vector of 390 distinct input lines ( axons arising from distinct neurons which may or may not be correlated in activity ) . Out of these 390 lines , a postsynaptic neuron may acquire a small fraction of this number , for example 20 input lines . Such an acquired set defines a subspace of the original space , and just as there is a 390-by-390 covariance matrix associated with the full input space , there is a specific 20-by-20 covariance matrix associated with the subspace defined by this neuron's input connections . Then for this neuron ( call it j ) , we can also associate a dominant eigenvector and its associated eigenvalue arising from j's covariance matrix . A simple theorem states that the weights of these inputs are proportional to this subspace’s dominant eigenvector . ( A pleasing result since this vector maximizes the information throughput compared to all other linear , n-by-1 input filters for a given y . ) Moreover , the theorem below tells us 1 ) the proportionality constant that scales this eigenvector to the stable weight values , 2 ) the average excitation of j , and 3 ) the variance of j's excitation . In what follows , we assume that ε is a small positive constant and assume synaptic modification has been going on for a long time with a fixed set of input lines . Therefore the synaptic weights , i . e . the column vector w ( t ) for neuron j , change very slowly and thus can be treated deterministically . For a fixed set of input lines , each input-activation is random column vector X ( t ) ( with realizations x ( t ) ) with mean value E[X] . Via the definition of excitation , y ( t ) = x ( t ) Tw ( t ) , the average excitation is E[Y] = E[XTw] . As noted above , the weights can be treated as a constant; thus in the limit , the mean excitation is E[Y] = E[XT]w ( ∞ ) . The variance of this excitation arises from the covariance matrix of the input to this neuron j . That is , define j’s covariance matrix of its input space as Cov ( X ) : = E[XXT] − E[X]E[XT] , and then note that w ( t ) TCov ( X ) w ( t ) =E[w ( t ) TXXTw ( t ) ]−E[w ( t ) TX] E[XTw ( t ) ] =E[Y2]−E[Y]2=Var ( Y ) ( 1 ) Finally , define λ1 to be the largest eigenvalue of this covariance matrix and e1 as its associated eigenvector of unit length ( the so-called dominant eigenvector ) . Theorem . Assuming a stable set of input weights is achieved via the synaptic modification equation Δwij = ε ( X ( t ) – E[X] – w ( t ) ) X ( t ) Tw ( t ) operating along with the shedding rule then , E[Y]=λ1w ( ∞ ) : =limt→∞w ( t ) =ke1 , wherek=Var ( Y ) /E[Y] . ( 2 ) Proof . By definition , stability implies limt→∞E[Δw ( t ) ]=0 . Then , taking this same expectation and limit on the other side of the synaptic modification equation yields E[Δw ( ∞ ) ] = 0 = ε ( ( E[XXT] − E[X]E[XT] ) w ( ∞ ) – w ( ∞ ) E[Y] ) , or ε ( Cov ( X ) w ( ∞ ) – w ( ∞ ) E[Y] ) = 0 , which implies Cov ( X ) w ( ∞ ) =w ( ∞ ) E[Y] . ( 3 ) Note that ( 3 ) is the eigen-equation , and the shedding rule guarantees all weights are positive while the synaptic modification equation guarantees wij is bounded from above because X–E[X] < 1 . Therefore because y is bounded both below and above , convergence is implied . With our old synaptic modification rule based on a correlation matrix of a non-negative input , the Perron-Frobenius ( PF ) theorem implies that the dominant eigenvector ( associated with λ1 ) is in the all-positive orthant . Here however , without an all-positive covariance matrix , we must conjecture an extension to PF ( see below ) . Thus , by this perturbation conjecture , the synaptic weights align with e1 ( proving 2 ) . Now designate an unknown positive constant k and define w ( ∞ ) = ke1 . Pre-multiplying ( 3 ) by e1T quickly yields ( 2 ) : e1TCov ( X ) w ( ∞ ) =e1Tw ( ∞ ) E[Y] implying λ1e1Tke1=e1Tke1E[Y] , producing the result E[Y] = λ1 . For ( 2 ) , pre-multiple both sides of ( 3 ) by w ( ∞ ) T , and note that by Eq ( 1 ) the left hand side is var ( Y ) while the right hand side yields k2E[Y] . Thus , k=Var ( Y ) /E[Y] If a neuron happens to acquire enough synapses , a valid central limit theorem ( with mean and variance of the excitation values ) would even tell us where threshold should be placed to produce the desired activity level . That is , the right-hand tail , beginning at threshold , yields the fraction of times a neuron fires to a randomly sampled input . This theorem assumes convergence of all algorithmic processes . However there is an important exception to the convergence hypothesis . Certain input configurations will never produce stable connectivities nor achieve their algorithmically guaranteed firing-rates . Sensibly , such neurons might be killed-off; such neurons might lower their firing threshold; or from another perspective , such an input configuration will be very unlikely to exist . For example , we must conjure an input environment in which a set of input patterns is orthogonal to all others ( thus very unlikely ) and the probability of a member of this set occurring is less than the receptivity cutoff . For example , suppose synaptogenesis remains positive until a neuron fires 10% of the time . Suppose a subset of patterns occurs 9% of the time and that this set of patterns is orthogonal to all the other patterns . If a subspace of this set with its positively correlated input lines gains a controlling influence on a postsynaptic neuron , then any other input line not positively correlated but acquired through synaptogenesis will have its weight decreased by the synaptic modification equation and then it will eventually be shed . Nevertheless , no matter how many positively correlated input lines are acquired , synaptogenesis continues never to halt ( because postsynaptic firing will converge to 9% , a value below the required 10% that halts synaptogenesis on such a neuron ) .
The theory of adaptive synaptogenesis was developed from observations of empirical neuroscience ( see [1] , [2] , and [16] for motivating studies ) , from the underlying assumption that in order for a neuron to be most useful , its afferent synapses must reflect the statistical structure of the input world , and from one more motivating idea . We assume that there are desirable operating values for balancing costs versus information ( e . g . mean firing-rate or mean excitation vs . variance of excitation ) . Then , as the outcome of the algorithm , adaptive synaptogenesis guarantees such desirable , predetermined values . In this regard , BCM theory led the way , as it explicitly creates postsynaptic neurons with a particular average excitation [5] . In this regard , BCM theory provided the inspiration for adaptive activity control over the long term . More generally , the importance of producing an average activity level in a postsynaptic neuron became clearer with the demonstration [17] that neuron parameters ( such a axonal leak current ) imply a particular optimum firing-rate in order to maximize the bits per joule of an axon . Given a neuron with such an optimized axon , the values of synaptic excitation must be important in terms of matching dendro-somatic-initial segment computation with the axon’s optimal firing-rate . As well , its synapses should in some sense maximize incoming information [15] . In any event , the BCM algorithm with initial full-connectivity conjoined with an appropriate shedding rule , may well produce identical results to what is found here . Of course spike-timing rules will also work , again assuming full initial connectivity [18] . Indeed , in its earliest version , the utility of adaptive synaptogenesis was understood in the context Barlow’s information-conserving compressive coding idea [19–20] , a clearly energy saving transformation with its reduction in both firing-rate and number of neurons while maintaining almost all of the information of the inputs . The idea of using random connectivity to create network codes has always been part of our synaptogenesis algorithm; in fact , it is the baseline condition in [6–7] . Independently , such ideas have been used to study efficient connectivity distributions [21] and abstract functions [22] . As documented in our early work [7] , just random connectivity without shedding is still quite useful for compressive coding . That is , these randomly formed networks produce large values of mutual information while decreasing statistical dependence . However , as documented in the series of articles [6 , 8–9] , random connectivity with associative modification is inferior to using the algorithm that includes synaptic shedding of small weights . Although we know of no first-principles theory for optimizing number of synapses , it is clear from synapse count data and the volume penalties incurred by synaptic structures [23] that only a minute fraction of an input space ( for example the lateral geniculate as the input to V1 ) can form synapses with any one postsynaptic neuron in the cerebral cortex . In this light , it may be possible to tune adaptive synaptogenesis to achieve an appropriate range of synapses per neuron . There are four differences between the adaptive synaptogenesis algorithm used previously and the current version: two of these ( A and B below ) are improvements that can be applied to the compressive algorithm , a third ( C ) is a specialization for neurons performing discrimination , and the fourth ( D ) is largely inconsequential in the context of the data structures used here . There are three primary results here: 1 ) extension of the adaptive synaptogenesis algorithm from data compression to discrimination; 2 ) documentation of neuronal allocation as a function of a category’s relative frequency and of the statistical input structure; and 3 ) when suitably formulated , adaptive synaptogenesis produces a stable connectivity in a stable input world . | One neural correlate of being an expert is more brain volume—and presumably more neurons and more synapses—devoted to processing the input patterns falling within one's field of expertise . As the number of neurons in the neocortex does not increase during the learning period that begins with novice abilities and ends with expert performance , neurons must be viewed as a scarce resource whose connections are adjusted to be more responsive to inputs within the field of expertise and less responsive to input patterns outside this field . To accomplish this enhanced , but localized improvement of representational capacity , the usual theory of associative synaptic modification is extended to include synaptogenesis ( formation of new synapses ) and synaptic shedding ( rejection of synapses by a postsynaptic neuron ) in a manner compatible with the original , associative synaptic modification algorithm . Using some mathematically simplifying assumptions , a theory is developed that predicts the algorithm's eventual outcome on expert neuronal coding , and then without the simplifying assumptions , computational simulations confirm the theory’s predictions in long , but finite periods of simulation-time ( i . e . , finite-sampling leads to stable connections , and thus , stable expert encodings ) . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Adaptive Synaptogenesis Constructs Neural Codes That Benefit Discrimination |
The ecto-5’-nucleotidase CD73 plays an important role in the production of immune-suppressive adenosine in tumor micro-environment , and has become a validated drug target in oncology . Indeed , the anticancer immune response involves extracellular ATP to block cell proliferation through T-cell activation . However , in the tumor micro-environment , two extracellular membrane-bound enzymes ( CD39 and CD73 ) are overexpressed and hydrolyze efficiently ATP into AMP then further into immune-suppressive adenosine . To circumvent the impact of CD73-generated adenosine , we applied an original bioinformatics approach to identify new allosteric inhibitors targeting the dimerization interface of CD73 , which should impair the large dynamic motions required for its enzymatic function . Several hit compounds issued from virtual screening campaigns showed a potent inhibition of recombinant CD73 with inhibition constants in the low micromolar range and exhibited a non-competitive inhibition mode . The structure-activity relationships studies indicated that several amino acid residues ( D366 , H456 , K471 , Y484 and E543 for polar interactions and G453-454 , I455 , H456 , L475 , V542 and G544 for hydrophobic contacts ) located at the dimerization interface are involved in the tight binding of hit compounds and likely contributed for their inhibitory activity . Overall , the gathered information will guide the upcoming lead optimization phase that may lead to potent and selective CD73 inhibitors , able to restore the anticancer immune response .
The immune response constitutes a major barrier for preventing cancer progression through the activation of T cells and subsequent release of pro-inflammatory cytokines . This process is initiated and tightly regulated by extracellular ATP which impacts a large variety of cells ( T and B lymphocytes , NK , macrophages , DC , neutrophils and vascular endothelial cells ) through the binding to P2X and P2Y receptors , inducing persistent inflammation and regulatory cell inhibition [1–3] . In healthy tissues , the extracellular ATP concentration is very low and estimated between 10 and 100 nM whereas in solid tumors , ATP is abundantly released in particular by dying cells , and through secretion , and its concentration can reach a few hundreds of micromolar [4] . In the tumor microenvironment , ATP usually acts as an alarm signal allowing the recruitment of immune cells and contributing to the immunogenic cell death process . However , when high ATP concentrations are associated with a high expression level of CD39 and CD73 on both immune and cancer cells , ATP is rapidly and successively degraded into AMP and then adenosine by the concerted activities of these two ectonucleotidases [5] . As a consequence , an abnormal adenosine concentration is produced in the tumor microenvironment and induces a potent suppression of the antitumor immune response through the adenosine binding to P1 receptors ( mainly A2a and A2b ) expressed on immune cells [6–9] . Ecto-5’-nucleotidase , or CD73 ( EC 3 . 1 . 3 . 5 ) , is a glycosylphosphatidylinositol ( GPI ) anchored cell surface protein that is expressed as a non-covalently linked homodimer on endothelial , immune and tumor cells . CD73 also exists as a soluble and circulating form with similar enzymatic activity to its membrane-attached form . Intriguingly , this soluble form was also found in cell and organ crude extracts probably generated by a phospholipase activity on the GPI-anchored precursor . However , the precise role of this intracellular form is not fully understood in particular because of the high intracellular ATP concentration making the enzyme inactive [10] . In human peripheral blood , CD73 is expressed on most of B lymphocytes , T cells including Th17 , NK and myeloid-derived suppressor cells [3] . These cells can also co-express CD39 and CD73 [11] . In the tumor microenvironment in which hypoxia is predominant , CD73 has been shown to be overexpressed in various types of solid tumors as well as endothelial cells [12] . This encompasses several cancers such as colorectal , breast , bladder , pancreas , ovarian , leukemia and melanoma , as recently reviewed in [13] , and is generally associated with poor prognosis in patients receiving anticancer treatments [14] . Few exceptions have been described pointing out CD73 as good prognosis marker as for the clinical study of endometrial and breast carcinomas [15 , 16] . This discrepancy between opposed roles of CD73 may be due to specific changes in endometrial cancers ( endometrial epithelial barrier integrity ) or may be a consequence of predominant presence of the soluble form of CD73 ( sCD73 ) . Indeed , higher plasma concentrations of sCD73 were determined in cancer patients or patients suffering of acute inflammatory pancreatitis compared to healthy individuals [17 , 18] . These studies suggest that the upregulation of sCD73 levels in blood may be a prognosis marker of tissue inflammation and tumor hypoxia . Moreover , CD73 overexpression has been shown to promote cell proliferation , migration , invasion and attachment to the extracellular matrix in human breast cancer [19 , 20] through the action of adenosine binding to A1 and A3 receptors [21] . CD73 deficiency was also studied in mice and correlated with resistance to in vivo development of carcinoma [22] or with antitumor immunity improvement [23] . The role of CD73-produced adenosine in cancer progression and metastasis has been evidenced by the use of either monoclonal antibodies [24] or siRNA [25] blocking CD73 enzyme activity . As a consequence , the immune response through ATP purinergic signaling could be restored . For all these reasons , CD73 has been considered as a promising therapeutic target to develop new anticancer therapies . The first described CD73 inhibitors were ADP , ATP and adenosine 5’-[α , β-methylene]diphosphate ( a non-hydrolysable ADP analog named APCP ) , all acting as competitive inhibitors [10] . Subsequently , small molecule inhibitors derived from APCP have been recently designed and studied showing potent competitive inhibition of CD73 [26] . However , competitive inhibitors , especially those targeting kinases , present several drawbacks such as low selectivity profiles [27 , 28] or weak efficiency when competing with high substrate concentrations . In order to overcome this problem , an alternative approach consists in the development of non-competitive or allosteric inhibitors interacting with the target outside the substrate binding site . As an evidence that occurred likely by chance , a new monoclonal antibody developed by MedImmune ( MEDI9447 ) , was shown to inhibit CD73 enzymatic activity through such a dual mechanism [29] involving a non-competitive inhibition . Although this approach was quite different in regard to small drug molecules , it demonstrates the proof of feasibility consisting in blocking the enzyme conformation and leading to CD73 inhibition . Interestingly , MEDI9447 did not compete with AMP and the antibody was able to prevent the conformational transition required for forming the enzyme active site . As illustrated by the crystal structures of this enzyme solved in the open and in the closed conformations [30 , 31] , large dynamics domain motions are obviously required to form the closed active conformation ( for both monomers ) . The di-metallic center is present in the N-domain while the adenosine moiety of the substrate binds to the C-domain and AMP hydrolysis will occur only after a large closure motion mediated by a rotation of the N-domain of up to 114° [30] . The objective of the current study was to reproduce the dynamics of the enzyme in order to identify druggable cavities and small molecule inhibitors able to block the dynamics and thereby the associated enzymatic function . To achieve this objective , we describe a bioinformatics approach that allowed the identification of new allosteric inhibitors ( designated hereafter as “RR” compounds ) targeting human CD73 and able to block efficiently its enzymatic function in the low micromolar range through a non-competitive inhibition mechanism . This is the first stage of a future drug development process and the selected lead compounds will require further structural optimization to envisage forthcoming in vivo applications . The final objective behind our search of new inhibitors is to restore the antitumor immune response by downregulating the extracellular adenosine concentration either by using RR compounds alone or in combination with immunotherapies .
The overall strategy followed in this study for cavity selection and hit identification is schematically illustrated in Fig 1A . First by analyzing the crystal structure ( 4H2G ) and by using the Fpocket program , we detected five potential druggable cavities ( Fig 1B ) . For the selection of the most suitable cavity , these pockets must fulfill important criteria: i ) a cavity located far away from the substrate binding site ( to avoid competitive inhibition ) , ii ) a cavity with a sufficient volume to afford the binding of drug-like molecules , iii ) a cavity showing variation in size and volume during the dynamics ( the final goal being to block dynamic motions of the enzyme ) , and iv ) displaying a high mean local hydrophobic density as previously described for druggability [32] . Therefore , to evaluate their change in size and volume during the reaction , we performed molecular dynamics simulations enabling the selection of the most suitable and druggable cavity for virtual screening . Hence , TMD simulations were carried out to reproduce the large domain motions occurring during the reaction in both directions ( from open to closed states and vice-versa ) . Indeed , to block the enzyme function , both directions are relevant as soon as the dynamic can be altered . We first focused on the closing direction in presence of the preferred substrate ( AMP ) . Rigid body motions of the N-terminal domain toward the C-domain were observed with a preponderant rotating motion during closure , all together leading to the formation of the active site ( Fig 1C and S1 Movie ) . Starting by an initial translational motion , both N-domains ( residues ranging from 27 to 337 ) operate an anti-symmetric rotation around a central node formed by four amino acids ( 335STQE338 ) located between α-helix I and β-sheet 15 [30] ( as shown in Fig 1C with the displacement of the center of mass of each N-domain ) . Similar results were obtained in the opposite direction except that large collective motions of the N-domain were slower at the beginning of the simulation ( using the same force constant applied in both TMD simulations ) . This can be easily explained by the presence of the substrate that is tightly bound to the C-domain through strong electrostatic interactions between the phosphate oxygen atoms and the two zinc ions . Using 100 conformers issued from the simulations ( Fig 1D ) , a cavity located at the junction of both C-domains ( called hereafter , dimerization interface , Fig 1E ) was the only pocket meeting all the druggability criteria previously defined , and was therefore selected as target binding site for virtual screening . As shown in Fig 1D , results obtained by MDpocket analyses indicated that the volume of this cavity is rather large and highly fluctuating ( from 2815 to 4732 Å3 ) and also that the mean local hydrophobic density was increasing for several conformers ( delineated by square symbols in Fig 1D ) . Therefore , five representative conformers ( denoted as C1 to C5 , S1 Fig ) were selected for an ensemble docking to mimic both the dynamics of the enzyme ( volume of the cavity ) and druggability according to the apolar descriptor . The center of the screening area was defined by both Glu543 residues located in this interface as shown in Fig 1F in the presence of one hit compound . The initial and final states of the simulation which correspond to the experimental crystal structures were not included for virtual screening as the previous defined criteria were not fully satisfied ( smaller volume of the pocket and apolar contribution ) . However , the overall structural quality of the selected conformers issued from the TMD simulation was assessed by means of three different methods . First , RMSD of backbone atoms from both C-domains were computed using as reference structure either the open or the closed crystal structure ( S1 Fig ) . Then , the Z-score ( Prosa protein analysis tool ) was calculated for each conformer and found to be very close to the ones computed for the experimental structures ( S2 Fig ) . In addition , the Ramachandran diagrams ( S3 Fig ) were computed for all structures and indicated a very low violation rate . Indeed , about 1% of residues were found in outlier regions and these latter were located near the substrate binding site . Altogether , these results indicated that the conformers selected for the virtual screening display an overall excellent structural quality . Virtual screening of 324 , 400 compounds was carried out by targeting the dimerization interface on the five conformers issued from TMD simulation ( ensemble docking ) . One could remark that half of the library was composed of compounds violating the Lipinski’s rules of 5 ( either by a molecular weight > 500 or by a clog P > 5 , or both ) . This feature was chosen on purpose for targeting the CD73 dimerization area as protein-protein interfaces are known to be highly apolar in comparison to exposed protein surfaces . The best hit compounds were selected from the top-ranked compounds obtained by AutoDock Vina and further rescored with Gold on each individual conformer . The final ranking was computed by averaging the score obtained with each conformer by Gold and a final round of selection was carried out to increase the structural diversity of hit compounds ( Fig 2 and S1 Table for the full list of selected hit compounds denoted as RR and ranked by docking score ) . A cut-off value of the computed docking score was arbitrary selected at 70 leading to a docking score range between 96 and 71 ( RR1 to RR28 ) . Additionally , five compounds ( fragment-like compounds ) from the initial library were kept for further testing because of their structural similarity with the best hit compounds ( five last molecules listed in S1 Table ) . Most of the hit compounds showed high clog P indicating that the targeted dimerization interface has indeed a large hydrophobic area . The 33 best-ranked compounds did not share a common chemical structure but present some interesting features like a 3-D shape exploiting the chemical space by combining rigid scaffolds such as five- or six membered aromatic rings either as a tri-branched based molecule often encountered as for compounds RR1-4 , 6 , 9 , 14 , 17–18 , 21 and 26 , or under an extended structure ( RR10-13 , 16 , 20 , 23–25 , 27 and 28 ) . Interestingly , four compounds , RR11 , RR13 , RR19 and RR28 are dimeric structures composed of two identical components linked together ( Fig 2 and S1 Table ) . This structural feature may be the indication of a common binding mode for each part of the compound with each monomer of the enzyme . Two molecules also exhibit a complex spatial organization as they include a ribose or a nucleoside scaffold bearing two or four aromatic protecting groups ( RR7 and RR8 , respectively ) . The selection of this type of multi-branched structure may arise from the large surface to be occupied in the binding site of the enzyme and consequently , this feature may contribute to the blockade of the protein dynamics or activity . In addition , two highly similar structures were both selected from the screening process , compounds RR4 and RR6 with comparable inhibitory activities . The tri-branched core is almost identical ( and based on a tetra-substituted pyridine ring ) except the nature of the substituent on the side chain located in position 2 and either corresponding to the 2-amino-4 , 5 , 6 , 7-tetrahydro-benzo ( b ) thiophene ( RR4 ) or to the ethyl 4-aminobenzoate ( RR6 ) . The end of the list shows some compounds with low scores , due to their low molecular weight comparable to fragments . These molecules were included for comparison and components analysis of larger molecules . We tested the 33 best-ranked compounds for their potential inhibition of CD73 activity using the recombinant purified enzyme ( S1 Table ) . For this purpose , the human dimeric soluble form of CD73 was expressed in insect cells using a pFastBac system to guaranty the presence of post-translational modifications since four potential glycosylation sites have been suggested [31] . Catalytic ( kcat ) and Michaelis ( KM ) constants were determined for the purified enzyme at 70 . 6 ± 2 . 4 s-1 and 4 . 8 ± 0 . 6 μM , respectively , leading to a catalytic efficiency of 14 . 7 μM-1 . s-1 . The enzyme was found three fold less active than the recombinant protein expressed in HEK cells [31] , but the activity was comparable to the commercially available human enzyme , also produced in HEK . As shown in Fig 3 , several compounds significantly inhibited the enzyme activity at a concentration as low as 5 μM . The most active ones , in terms of inhibition of CD73 enzyme activity , were RR2-4 , 6 , 8–9 , 11 , 16 , 18 and 20–21 which promoted an enzyme inhibition with a similar efficacy to that observed with APCP ( Fig 3 ) . Higher concentrations of RR compounds ( up to 200 μM ) were tested giving a similar inhibition profile with larger standard errors due to the poor water-solubility of these compounds . Interestingly , some compounds gave negative values of inhibition meaning that they were able to activate the enzyme . This result was not surprising since allosteric compounds may play the opposite role by stabilizing a preferential conformation leading to higher enzymatic efficiency ( positive allosteric regulators ) . RR28 was the remarkable example of this type of enzyme enhancers and also RR12 and RR14 in a lesser extent . The strongest inhibitor was compound RR3 that induced 93% of enzyme inhibition at a concentration of 5 μM . It was also predicted as the less water-soluble compound ( clog P value: 8 . 7 ) . Since these compounds were highly hydrophobic , we computed several metrics commonly used in drug design such as LE ( ligand efficiency ) , LLE ( ligand-lipophilicity efficiency ) , BEI ( binding efficiency index ) and SEI ( surface-binding efficiency ) ( Fig 4 and S1 Table ) in order to better evaluate the physicochemical properties that are preponderant in the binding efficiency [33 , 34] . LE is a simple but important indicator to select compounds according to their efficacy in respect to their atom number count . For orally available active compounds compliant with the rule of five , LE value should be at least 0 . 3 and this value is used for the selection of leads and needs to be maintained during the optimization process [35] . Here , all hit compounds exhibited low values of LE between 0 . 16 and 0 . 23 kcal . mol-1 . HA-1 with RR3 as best lead ( LE = 0 . 23 with pKi = 6 . 28 ) followed by RR6 , 9 , 11 , 16 , 18 and 20 ( LE = 0 . 22 ) . In contrast , LLE , which takes into account lipophilicity , indicated that the lipophilic contribution for RR3 was not optimal ( LLE = -2 . 48 ) and in this respect , RR16 appeared as the better compound ( LLE = 1 . 97 and LE = 0 . 22 with pKi = 6 . 77 ) and finally RR4 and RR6 with a moderate lipophilic contribution . Therefore , BEI and SEI were also computed to better appreciate which compound involves its molecular structure in the binding or the inhibition efficiency . While BEI takes into account only the molecular weight ( global size ) , SEI encompasses the polar surface area ( between 44 and 160 Å2 ) , reflecting much better the occupation efficiency of the molecular surface . Here , RR3 showed the highest value followed by RR20 . Because of the structural shape analogy between RR10 ( weak inhibitor ) and RR16 ( strong inhibitor ) , the kinetic inhibition assay was also carried out for RR10 . Indeed , the determined Ki value ( 9 . 4 μM ) was higher than the respective ones for RR16 ( 0 . 46 and 1 . 7 μM , mixed inhibition ) . Although the inhibition mode was different for both compounds , the calculated LLE value was much higher for RR16 than for RR10 but these compounds showed very similar BEI or SEI values . According to these indexes , we can conclude that: RR3 should be improved for a better use of its lipophilicity ( LLE too small ) as for RR20; RR8 has very low pKi , LE and LLE values render difficult its optimization and finally , RR16 could be improved for a better use of its molecular surface . Due to the low water solubility , this analysis could not be performed for compound RR21 . According to the location of the target binding site that was far away from the substrate binding site , RR compounds should impair the enzymatic function through a non-competitive inhibition mode . As expected , the kinetic mechanism describing the inhibitory activity for eight compounds ( RR3 , 4 , 6 , 8–11 and 20 ) was the non-competitive mode ( Fig 5 and S1 Table ) . Nevertheless , for some compounds like RR2 and RR16 , a mixed inhibition mode was determined , indicating that they may also bind to the substrate binding site or to the enzyme-substrate complex . The inhibition profile could not be determined for poorly water-soluble compounds like RR21 . Also , the inhibition mechanism could not be defined unambiguously for compound RR18 , for which the experimental data fitted well with both mixed and non-competitive equations . The most active non-competitive inhibitors were RR3 and RR6 with Ki values of 0 . 52 ± 0 . 20 μM and 0 . 68 ± 0 . 05 μM , respectively ( Fig 5A and 5C ) . RR4 and RR20 were less potent than the previous ones but still able to induce a strong inhibition of CD73 activity and exhibited a Ki around 1 . 2 μM ( Fig 5B and 5D ) . RR16 was deduced as a potent mixed inhibitor of CD73 with Ki and Ki’ values of 0 . 46 ± 0 . 10 and 1 . 70 ± 0 . 20 μM , respectively . As shown in Fig 6A , selected hit compounds were predicted to bind entirely to the large targeted cavity and they were spanning at least three sub-parts of the cavity ( Fig 6B ) . Focusing on the most active compounds ( RR3 , RR6 and RR16 ) all of them were deeply buried in the dimer interface and all three hits interact with one or two glutamate residues ( E543 ) . However , their binding modes were found to be slightly different . Indeed , for each compound the main interactions with amino acid residues were different , I455 and E543 with RR3 ( Fig 6C ) , K471 Y484 and E543 ( backbone oxygen from both residues from the two monomers ) with RR6 ( Fig 6D ) , D366 , I455 , H456 , Y484 ( both ) and E543 ( both ) with RR16 ( Fig 6E ) . In addition , a halogen bond is formed between V542 backbone oxygen and the chlorine atom of RR16 . Interestingly , RR16 was connected to a huge number of amino acid residues in contrast to the other hits . The presence of two sulfone groups may explain this distinct binding . On the other hand , several hit compounds contained a stretched or more spanned chemical structure like RR11 or RR20 and were determined as weaker inhibitor than RR3 . Consequently , a rigid structure may be unfavorable for a tight binding . In addition to its rigidity , RR11 exhibits a dimeric structure and binds to CD73 with a different orientation compared to RR3 even though it was found deeply inserted into the dimer interface ( Figs 6F and 7C ) and making two polar interactions with D366 and Y484 . Since the virtual screening was achieved on five conformers and the docking analysis done on an unique conformer ( the main selected one by the ensemble docking and leading to highest computed scores ) , we analyzed the variations in binding mode for the most interesting compounds depending on the conformer used for docking ( S4 Fig ) . As shown in S4 Fig , slight variations ( average RMSD of 0 . 5 Å ) were observed except for RR3 for which the RMSD was between 1 . 2 and 1 . 8 Å . Also , the docking onto conformer C1 often led to a distinct binding in comparison with other conformers . However , the binding mode was found very similar between at least two conformers over the five for the large majority of hit compounds as for RR4 , RR6 , RR9 and RR16 for instance . The conformers leading to the best scores were C3 and C5 corresponding to the middle and the end of the simulation ( from open to closed states ) . The preferential binding to conformer C5 may be explained by its larger cavity volume and higher hydrophobicity than for C3 suggesting that these properties allowed to afford a stronger binding . It also shows that the binding may be more efficient when the enzyme is closing to form its active site in presence of the substrate . This comparison also indicates that the large collective motion of the enzyme modifies substantially the target cavity in terms of steric space highlighting the importance of using multi-conformational states during the virtual screening . A detailed analysis of the docking binding poses indicates that the hydrophobic contribution in the binding efficiency of hit compounds was quite important as predicted by clog P values . Indeed , in addition to the most encountered residues making hydrogen bonds with RR compounds ( D366 , H456 , K471 , Y484 and E543 ) numerous apolar residues participated in hydrophobic contacts such as G453-454 , I455 , L475 , V542 and G544 . Also , two polar residues , H456 and Y484 are also involved ( Figs 7 and 8 ) . Moreover , a few charged residues contributed to these non-bonded interactions such as D366 , K471 , D473 , E543 and R545 reinforcing likely the binding affinity . We first compared the two structurally related compounds RR4 and RR6 to highlight their binding mode , orientation and differences ( Fig 7A and 7B ) . These molecules were almost superimposable in their binding site . Highlighting their differences , we observed that the methoxyphenyl group interacts with a patch of glycine residues ( G453-G454 ) for RR6 and this was not seen with RR4 . Moreover , the van der Waals contacts involved Y484 and V542 for RR4 while it was replaced by F548 in the case of RR6 . This little difference observed in the predicted binding modes may explain the two-fold factor between Ki values for these two compounds ( inhibition constants very close to each other ) . For elongated and more rigid structures like RR11 and RR20 ( Fig 7C and 7D ) , no binding similarities could be observed as one molecule is curved while the other is more stretched allowing to cover a larger surface area in the binding site . This may be explained by their different degree of rigidity . Nevertheless , both compounds connect the two monomers together leading to an enzymatic inhibition . This is achieved through numerous hydrophobic contacts as shown by the interaction with apolar residues . According to their respective Ki values , a flexible chemical structure seems to be less advantageous for the inhibition efficiency , most probably because of the entropic loss upon binding to CD73 . Interestingly , RR28 was found to increase the enzyme activity instead of impairing it . To understand the reasons why this molecule to behave as an allosteric activator , we compared its binding to the most potent non-competitive inhibitor , RR3 . As shown in Fig 7E and 7F , RR28 did not bind to CD73 in the same orientation and less hydrophobic contacts were found . Two residues ( D366 and Y484 ) are making halogen bonds by interacting with the fluorine atoms . However , apolar residues ( I369 and L465 ) are very close to the fluorine atoms leading to unfavorable contacts . This may explain why the docking score obtained for RR28 was much lower than that of RR3 ( 71 versus 94 ) and suggests a weak binding in this pocket . Moreover , the structure of this compound shows an axial symmetry enabling to link both monomers of the enzyme through halogen and hydrogen bonds ( Fig 8 ) . Therefore , one can imagine that the dimer is better stabilized in the presence of RR28 . This may explain why RR28 was found to act as an enhancer of CD73 activity but another explanation would be that it binds to another site to promote such unexpected effect . Nevertheless , two identical residues ( D366 and Y539 ) from each monomer are connecting the hit compound suggesting that this symmetrical compound takes benefit of the symmetry of the dimer .
From this virtual screening study targeting the dimerization interface of CD73 as potential allosteric binding site , several hit compounds were determined as strong non-competitive inhibitors and other as mixed inhibitors . The most active compounds exhibited Ki values in the low micromolar range allowing for further hit to lead optimization . Structure of hit compounds were characterized by two different scaffolds either as a tripartite shape or an extended structure with similarities within the two families . The large structure allowed spanning the large cavity . Interestingly , some compounds were shown to create a strong linkage between the two monomers leading either to an inhibition or an enhancement of the enzymatic activity . Allosteric regulation has been extensively described for many enzymes , especially kinases like p21-activiated kinase 4 [36] , small GTPases [37] or G-protein-coupled receptors for more than half a century [38 , 39] . All these proteins or enzymes play an important role in maintaining the cell integrity or signaling and have also been pointed out for therapeutic approaches including the development of new cancer treatments . Nowadays , the design of allosteric compounds represent a valuable alternative approach to identify new drugs targeting proteins that are considered “undruggable” by developing either positive or negative allosteric modulators [40] . Within the large Halo-Acid Dehalogenase family from which CD73 belongs , the cytosolic 5’-nucleotidase II is a good example of allosteric regulation by ATP or bisphosphoglycerate as previously described [41 , 42] . Here , we targeted an interface that is not described as an allosteric site . However , an allosteric activation was observed with RR28 in addition to the strong inhibition induced by several hit compounds , indicating that the target binding site was able to modulate the enzymatic activity through the binding of small molecules . It must be highlighted here , that the data obtained so far do not allow us to conclude that inhibitors bind actually in the assumed allosteric binding site ( targeted during the virtual screening ) and this conclusion will only become definitive by solving the crystal structure of the complex , for instance . Similarly , it cannot be excluded that RR28 binds to a different allosteric pocket to that of RR3 or RR4 . Also , another question remains concerning the allosteric effect that includes by definition , a protein conformational change . Here , we could not measure experimentally this effect and we assume that non-competitive inhibitors act as allosteric inhibitors ( and the opposite for RR28 acting as allosteric activator ) . The chemical nature of the identified compounds leads to high lipophilicity according to their clog P values and consequently a lower water solubility . This arises from the selection of hydrophobic compounds present in the chemical library during the screening phase when targeting the interfacial binding site . Screening in the substrate binding site would have selected more hydrophilic compounds . Nevertheless , the current hit compounds will have to be optimized to increase their bioavailability . This step can be achieved by several methods often used in all drug discovery programs ( search by similarity or pharmacophore models ) and keeping in mind that a certain degree of lipophilicity is required to ensure a tight binding in the dimer interface . Alternatively , permeation enhancers may be useful to improve their physicochemical properties before reaching the enzyme target such as cyclodextrin-based formulations [43–45] or by using chitosan [46] or glyconucleolipid [47 , 48] derivatives leading to both an increase in bioavailability and half-life of the compound . The main objective of this study was to block the enzymatic activity by hindering the dynamics of the enzyme that is required for its function ( and therefore the active site formation leading to the hydrolysis of AMP into adenosine ) . One interesting feature here is the selected cavity that is located at the dimer interface and can be therefore considered as a protein-protein interface . This point was already discussed in previous publications targeting protein-protein or protein-DNA interfaces and led to the discovery of interfacial inhibitors , like Brefeldin A binding to the Arf-Sec7 interface or camptothecin binding to topoisomerase I-DNA complex [49] . CD73 has been extensively studied for its implication in cancer development and progression [4] and in addition to the monoclonal antibody ( MEDI9447 ) [29] , several small molecule inhibitors have been developed . All these compounds ( anthraquinone [50] , sulfonic acid or sulfonamide derivatives [51 , 52] or those being derived from APCP [26] ) were designed by targeting the substrate binding site or by analogy to the substrate itself , and most of them act as competitive inhibitors . One exception has been recently described with 2-alkoxy-3- ( sulfonylarylaminomethylene ) -chroman-4-one derivatives acting as uncompetitive inhibitors [53] . This study indicates that these inhibitors block the enzyme by targeting an enzyme-substrate intermediate of the reaction . This was the first suggestion of the presence of a binding site different to that of the substrate . Here , we describe for the first time the inhibition of CD73 activity by a most likely allosteric mechanism , which may lead to higher enzyme selectivity and less off-target effects .
The overall strategy is schematically illustrated in Fig 1A . Targeted molecular dynamics ( TMD ) simulations were carried out using two crystal structures of CD73 in the direction from the open ( 4H2G ) to the closed ( 4H2I ) conformation . All calculations were performed with NAMD 2 . 11 [54] in the isobaric–isothermal ensemble . The pressure ( 1 atm ) and temperature ( 310 K ) were kept constant using Langevin dynamics and Nosé-Hoover Langevin piston [55 , 56] . All protein atoms and Zn ions were described by the CHARMM27 force field [57] . The substrate AMP was modelled using adenosine structure from 4H2G and inorganic phosphate from 4H1S by structural alignments of respective C-domains . Missing parameters in Charmm force field were added by homology to ADP but with atomic partial charges computed with Gaussian ( RHF/6-31G ) by fitting the electrostatic potential surface . The system was solvated with explicit water ( TIP3P model ) , neutralized with four sodium ions and replicated in each direction using periodic boundary conditions . The short-range Lennard-Jones potential was smoothly truncated from 10 to 12 Å and the PME ( Particle Mesh Ewald ) algorithm [58] was used to calculate long-range electrostatics with a grid spacing of 1 Å . The potential energy of the molecular systems was minimized for 100 , 000 steps of conjugate gradient ( time step of 2 fs ) . After a gradual heating from 0 to 310 K , the two systems were further equilibrated for 100 , 000 steps . A spring force constant of 200 kcal/mol/A2 was applied to all atoms and defined in the TMD potential term ( UTMD , see Eq 1 ) allowing reducing the root mean square ( RMS ) distance between open ( 4H2G ) and closed ( 4H2I ) conformations during 20 ns . The two C-domains ( residues 337–549 ) of both structures were aligned prior to simulation . UTMD=12kN[ RMS ( t ) −RMS* ( t ) ]2 ( Eq 1 ) where RMS ( t ) is the instantaneous best-fit RMS distance of the current coordinates from the target coordinates , and RMS* ( t ) evolves linearly from the initial RMSD at the first TMD step to the final RMSD at the last TMD step . The elastic constant k is scaled down by the number N of targeted atoms . Identification and characterization of druggable cavities were achieved with Fpocket or MDpocket software [59] . Selection criteria were the volume and mean local hydrophobic density ( ratio of neighboring apolar alpha spheres divided by the total number of apolar alpha spheres in the pocket ) ; this ratio is then normalized in respect to the other binding pockets [60] . The potential energy function of the five conformers selected from TMD simulation ( and further used for ensemble virtual screening ) was minimized with 50 , 000 steps of conjugate gradient using NAMD . The conformers were subjected to a careful structural quality assessment using the ProSA-web server [61] ( https://prosa . services . came . sbg . ac . at/prosa . php ) ( S2 Fig ) and by computing their Ramachandran diagrams ( S3 Fig ) by using the Rampage program hosted at the University of Cambridge [62] in order to compare the overall quality with the experimental crystal structure ( 4H2G and 4H1S ) . A chemical library of 324 , 400 compounds was generated from the Molport screening compound database gathering 34 suppliers and composed of natural and synthetic molecules with drug-like properties ( http://www . molport . com ) . The library of screening compounds was composed of unique molecules , commercially available from several main suppliers ( Asinex , ChemDiv Inc . , Vitas-M laboratory and Enamine ) . Before using it for virtual screening , the library was filtered in order to remove duplicates , add explicit hydrogens , generate 3D coordinates and finally to transform in PDBQT ( Vina ) or Sybyl mol2 ( Gold ) format using Open Babel 2 . 4 . 1 [63] . Despite a careful filtering , few compounds escaped to the modified Lipinski’s rules of 5 ( initial Ro5 with a molecular weight allowed to be greater than 500 Da and a clog P greater than 5 ) as shown by some low molecular weight fragments found in the library . Virtual screening was performed using the highly parallelized implementation , VinaLC-1 . 1 . 2 [64] of the Autodock Vina molecular docking program [65] . A confirmation of the docking poses was achieved using a second program and scoring function ( GOLD 5 . 2 program , CCDC Software Limited , [66] ) in order to increase the prediction accuracy . The center of mass of the two E543 residues ( from both monomers ) , located in the vicinity of the interface was targeted with radius of 15 Å around this point . The goldscore function was used to rank the docking solutions by using the clustering method ( complete linkage ) from the RMSD matrix of solutions . As the conformer C3 ( S1 Fig to S3 Fig ) was selected most of the time during the ensemble docking with the 5 TMD conformers and led to the highest scores when tested separately , this conformer was kept for the docking analysis of all hit compounds ( SAR relationships ) . Molecular dynamics simulations were analyzed with the VMD software [67] and structural analysis and visualization of docking poses were prepared using the PyMOL Molecular Graphics System ( version 1 . 8 , Schrödinger , LLC ) . For molecular interactions between CD73 and RR compounds , a maximum cutoff distance of 3 . 5 Å and 4 . 5 Å was used for hydrogen bonds and van der Waals contacts calculations , respectively . The clog P values for RR compounds were calculated using the robust Molinspiration chemoinformatics utility and the mi-log P model ( www . molinspiration . com ) . Various ligand efficiency metrics have been computed such as LE for ligand efficiency ( LE = [1 . 4 x ( -log Ki ) ]/NHA , where NHA is the number of heavy atoms excluding hydrogens and expressed in kcal . mol-1 . HA-1 ) [34 , 68] , LLE or ligand-lipophilicity efficiency ( LLE = pKi−cLog P , where pKi = -log ( Ki ) ) , BEI or binding efficiency index ( BEI = pKi / molecular weight in kDa ) , SEI or surface-binding efficiency index ( SEI = ( pKi ) / ( Polar surface area /100 Å ) [69] . The plasmid with NT5E gene encoding for the human soluble sCD73 protein ( residues 27–549 ) was kindly provided by Prof . N . Scrutton [31] . This construct already contained a His-tag at the C-terminus and a signal sequence derived from human extracellular glycoprotein ( osteonectin , residues 1–19 ) followed by Leu-Ala-Ser allowing extracellular expression of sCD73 [70] . The insert was subcloned into pFastBacTM vector 1 ( ThermoFisher Scientific ) after PCR amplification to include EcoRI/NotI restriction sites . Protein was expressed in Sf9 insect cells ( Life Technologies ) using the pFastBac baculovirus system ( ThermoFisher Scientific ) according to the manufacturer’s instructions . Insect cells were grown in suspension with stirring at 110 rpm in EX-CELL 420 medium at 28°C ( Sigma ) up to a density of 4×106cells per mL and then infected with baculovirus encoding CD73 . The cellular supernatant was harvested by centrifugation ( 20 min/31 , 000 g ) 48 h post-infection , filtered ( 0 . 22 μm ) , supplemented with protease inhibitors ( leupeptin , benzamidine and PMSF at 100 μg/mL ) and concentrated on crossflow cassette ( Vivaflow 200 Sartorius ) . The concentrate was centrifuged ( 30 min/186 , 000 g ) and purified on HisTrap Excel column connected to a FPLC Äkta purifier system ( GE Healthcare Life Sciences ) . The enzyme purity , size and activity were assessed by SDS-PAGE , Western blot and steady-state kinetics with various AMP concentrations . An extinction coefficient of 56 , 310 l . mol-1 . cm-1 was used for determining protein concentration at 280 nm . The 33 hit compounds were purchased from MolPort compound order service ( www . molport . com ) gathering all compounds from various suppliers . The purity and structural integrity of the purchased chemical compounds have been evaluated by NMR and mass spectroscopy ( S5 Fig ) . Adenosine 5′-monophosphate sodium salt ( AMP ) was purchased from Sigma-Aldrich and adenosine 5’- ( α , β-methylene ) diphosphate ( APCP ) used as positive control was synthesized using a previously published procedure [26] . The CD73 nucleotidase activity was determined by steady-state kinetics measuring the adenosine produced upon AMP hydrolysis by CD73 over time . The reaction was carried out in a thermostatically controlled beaker under magnetic stirring at 37°C in a buffer containing Tris-HCl 50 mM pH 7 , NaCl 100 mM , MgCl2 1 mM , CaCl2 1 mM . Reaction was allowed to occur upon addition of the substrate and stopped by addition of 10% of perchloric acid every 5 s . The same procedure was repeated in presence of each inhibitor at 5 μM . Reaction products were quantified by HPLC chromatography ( Waters Alliance ) using a Partisphere 5-SAX column ( AIT France ) and 10 mM ammonium phosphate buffer pH 5 . 5 as mobile phase . For non-water soluble compounds , DMSO was used and the final percentage did not exceed 0 . 5% in order to preserve the full enzyme activity ( enzyme tolerance for DMSO was determined up to 2% ) . The commercial human CD73 enzyme produced in eukaryotic cells ( Interchim ) was used to compare the kinetics parameters of both batches and to confirm the inhibition promoted by hit compounds . For the most interesting compounds , the inhibition mode was determined by steady state kinetic assays to obtain apparent catalytic ( kcat ) , Michaelis ( KM ) and inhibition ( Ki ) constants . Recombinant enzyme ( 2 . 5 nM ) , substrate ( AMP , at eight different concentrations: 1 , 1 . 5 , 2 . 5 , 5 , 10 , 25 , 50 and 100 μM ) were mixed in a thermostated beaker at 37°C in the presence or in the absence of inhibitors and reaction was stopped every 5 s by acid quenching before HPLC analysis ( as mentioned above ) . Quantification of adenosine and AMP was achieved by integrating peaks ( Empower software , Waters ) and raw data were analyzed using Grafit 7 ( Erithacus software ) and fitted with four different equations describing either a competitive , uncompetitive , non-competitive or mixed inhibition mode . The best model ( with the lowest Chi square value ) fitting the experimental data was considered as the inhibition mode and used for determining Ki . All experiments were carried out using three different inhibitor concentrations and Lineweaver-Burk plots were drawn to illustrate the inhibition modes . | Nucleotidases play a central role in maintaining the nucleotide pool homeostasis and the only extracellular member of this family , CD73 , has become an attractive target in oncology because of its high expression level on immune and cancer cells . In the tumor microenvironment , CD73-generated adenosine prevents the pro-inflammatory response and is considered as a potent immune suppressor . The current study aimed at developing new CD73 inhibitors by targeting an allosteric binding site in order to block the enzyme dynamics and therefore its enzymatic function . Most of the existing inhibitors have been elaborated on the basis of the substrate skeleton and act as competitive inhibitors . Here , four non-competitive compounds are presented with an inhibition constant in the low micromolar range . This study confirms the existence of an allosteric binding site located at the dimerization interface allowing modulation of the enzyme activity by small molecules , similarly to a previously described monoclonal antibody . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"glycosylamines",
"medicine",
"and",
"health",
"sciences",
"crystal",
"structure",
"enzymes",
"immunology",
"condensed",
"matter",
"physics",
"enzymology",
"molecular",
"biology",
"techniques",
"crystallography",
"enzyme",
"inhibitors",
"adenosine",
"physical",
"chemistry",... | 2018 | Identification of allosteric inhibitors of the ecto-5'-nucleotidase (CD73) targeting the dimer interface |
Differentiated cells can be reprogrammed through the formation of heterokaryons and hybrid cells when fused with embryonic stem ( ES ) cells . Here , we provide evidence that conversion of human B-lymphocytes towards a multipotent state is initiated much more rapidly than previously thought , occurring in transient heterokaryons before nuclear fusion and cell division . Interestingly , reprogramming of human lymphocytes by mouse ES cells elicits the expression of a human ES-specific gene profile , in which markers of human ES cells are expressed ( hSSEA4 , hFGF receptors and ligands ) , but markers that are specific to mouse ES cells are not ( e . g . , Bmp4 and LIF receptor ) . Using genetically engineered mouse ES cells , we demonstrate that successful reprogramming of human lymphocytes is independent of Sox2 , a factor thought to be required for induced pluripotent stem ( iPS ) cells . In contrast , there is a distinct requirement for Oct4 in the establishment but not the maintenance of the reprogrammed state . Experimental heterokaryons , therefore , offer a powerful approach to trace the contribution of individual factors to the reprogramming of human somatic cells towards a multipotent state .
Reprogramming somatic cells to become ES-like is an important goal in cell replacement therapy since it affords the opportunity to generate and use patient-specific ES derived cells as grafts . Epigenetic reprogramming can be achieved in different ways including nuclear transfer [1]–[4] or the forced expression of one or more transcription factors [5] , [6] . Retroviral-mediated expression of four transcriptional regulators , Oct4 , Sox2 , c-Myc and Klf4 , was shown to drive mouse fibroblasts into an ES-like ( iPS ) state , albeit at low frequency [7] , [8] . Reprogramming of human fibroblasts has also recently been achieved in a parallel approach using Oct4 , Sox2 and either Nanog plus Lin28 [9] or Klf4 plus c-Myc [10] . These pioneering studies have illustrated the importance of several factors for iPS , but also suggested that additional ones may be needed for efficient conversion to pluripotency . Reprogramming can also be achieved by cellular fusion , a process that occurs spontaneously in vitro [11] , in vivo [12] and experimentally using specific agents [13] . For example , fusion of differentiated cells with pluripotent ES cells , embryonic carcinoma ( EC ) or embryonic germ ( EG ) cells , induces the expression of pluripotency-associated markers in the hybrid cells [14]–[18] and chromatin remodelling at specific sites in the somatic cell genome [14] , [15] , [18]–[21] . While these data show that reprogramming occurs through the epigenetic resetting of gene expression programs in the differentiated cell , it has been unclear whether nuclear fusion and genome duplication are absolutely required for successful conversion [22] . Here we investigated the requirements for , and the stability of , dominant reprogramming of human B cells by fusion with mouse ES cells . We show that reprogramming is surprisingly rapid and occurs within heterokaryons in which lymphocyte and ES cell nuclei remain spatially discrete . Furthermore , our data show that while Oct4 is critical for successful reprogramming of human lymphocytes to an ES-like state , Sox2 is not required . Thus our data outline an alternative strategy for defining the factors that are required for inducing a pluripotent state in human somatic cells .
Human B cells were fused with mouse ES cells using polyethylene glycol ( PEG ) . The nuclear events in fused cells were monitored by fluorescence microscopy and quantitative RT-PCR to analyse gene expression ( Figure 1 ) . To facilitate the identification of fused cells , E14tg2a mouse ES cells were pre-labelled with DiD and human B cells with DiI and dual-stained cells were purified by FACS ( typically 10–15% of cells , Figure S1A ) . Human ( B cell-derived ) and mouse ( ES cell-derived ) nuclei were distinguished on the basis of DAPI and human-specific Lamin A/C labelling , and the proportion of cells containing two discrete ( heterokaryons ) or conjoined nuclei ( hybrids ) was assessed over time ( Figure 1B ) . Up to 2 days following cell fusion 98–99% of dual labelled cells were identified as heterokaryons in which a single human and a single mouse nucleus were evident ( illustrated in Figure 1B , central image ) . The kinetics of nuclear fusion were also confirmed by fluorescence in situ hybridization ( FISH ) analysis in which probes specific for mouse chromosomes ( γ-satellite , red ) or human chromosomes ( α-satellite , green ) were used to detect interspecies chromosome mixing , indicative of hybrid formation ( Figure S1B and Text S1 ) . The expression of pluripotency-associated genes and lymphocyte-associated genes by human B cell-derived nuclei was assessed by qRT-PCR , using primers that selectively amplify the human transcripts . Expression of human Oct4 , Nanog , Cripto , Dnmt3b and Tle1 was detected in cells as early as 1 day after fusion and human Rex1 after 2 days ( Figure 1C and Figure S1C ) . Although the levels were low in heterokaryons ( <1% of that detected in human ES cells , cell line NCL1 ) , these increased over time and were undetectable in non-fused ( or self-fused , not shown ) human B cells or control mouse ES cells . Expression of hTert was detected from day 4 onwards ( Figure S1D ) , while hHprt expression was equivalent at all stages , as anticipated ( Figure 1D ) . Mouse lymphocyte-specific gene transcripts ( mCD19 , mCD37 and mCD45 ) were not detected throughout the analysis ( not shown ) , confirming the dominance of ES cells in conversion [15] , [18] . Increased expression of human pluripotency-associated genes over this 8-day period was mirrored by a reduction in expression of several human lymphocyte-associated genes within the second ( hCD45 , hCD37 and hCD19 ) or third day ( hCD20 and hPax5 ) of heterokaryon formation ( Figure 1D ) . Collectively these data show that upon dominant reprogramming , activation and silencing of tissue-specific gene programs begins ahead of , and therefore does not require , nuclear fusion and cell division . In addition , since these results examine gene expression at the population level , it is possible that gene expression varied between individual heterokaryons and hybrid cells . As the reprogramming of somatic cells has been previously shown to result in altered DNA methylation at specific loci [15] , [18] , [21] , [23] , we examined changes in the methylation status of the human Oct4 gene promoter [24] and as a control , the Igf2/H19 imprinting control region ( ICR ) [25] . As illustrated in Figure 1E , human B cells prior to fusion showed high levels of DNA methylation throughout the hOct4 promoter and across a single Igf2/H19 allele . Following cell fusion , DNA methylation of hOct4 in reprogrammed B cells declined , consistent with a trend towards a hypomethylated state as seen in the human ES cell line H1 . Demethylation of the hOct4 promoter was detected prior to nuclear fusion and cell division , a result that is consistent with active chromatin remodelling of the locus prior to expression . No changes in DNA methylation at Igf2/H19 ICR were detected over this period , consistent with its imprinted status [25] . A comparison of the relative abundance of gene-specific transcripts in reprogrammed human B cells ( Figure 2A , right-hand column ) , showed a strong similarity with the gene expression profiles of several human ES cell lines ( NCL1 [26] , HI , H7 , H9 [27]; Figure 2A , left-hand column ) . For example , while Oct4 was abundantly expressed in all human and mouse ES cell lines , Nanog and Cripto expression was consistently much lower than Oct4 ( 100–1000 fold ) for each of the mouse ES cell lines analysed ( OS25 , CCE , E14 , ZHBTc4; Figure 2A , middle panel ) . In human ES cell lines however , Oct4 , Nanog and Cripto transcripts were similarly abundant , consistent with that seen in reprogrammed human B cells . Expression of some pluripotency-associated genes , for example Sox2 , was variable and often required extended periods of time ( >8 days ) for detection ( not shown ) . This could reflect the fact that genes such as Sox2 are subject to multiple layers of repressive epigenetic modifications in B cells including DNA and histone methylation [28] , [29] and late replication [30] , or that they require a higher threshold of activators for overt expression . Similarities between gene expression profiles of human ES cell lines and hB x mES fused cells prompted us to examine additional markers that are expressed solely by either human or mouse ES cells [31]–[34] . These included fibroblast growth factor receptors ( Fgfr1 and Fgfr2 ) and Fgf2 ( expressed by human ES cells ) , Bmp4 and leukaemia inhibitory factor ( Lif ) receptor ( expressed by mouse ES cells ) and SSEA4 , a surface glycoprotein selectively expressed by human ES cells [27] ( Figure S2B ) . This analysis revealed that reprogrammed cells expressed increasing amounts hFgfr1 , hFgfr2 and hFgf2 but did not express hBmp4 or hLifr or upregulate the downstream kinase hJak3 ( Figure 2B ) . Thus , these data show that while dominant conversion is driven by mouse ES cells ( that express Bmp4 and Lifr prior to fusion , Figure S2A ) , reprogrammed heterokaryons and hybrid cells show a remarkably different expression profile resembling human , rather than mouse ES cell lines . Consistent with this , fusion of mouse ES cells and human B cells resulted in SSEA4 expression by 13–16% of the cells ( days 2–8 as shown in Figure 2C ) . Isolation of SSEA4-positive cells confirmed that this subset contained successfully reprogrammed cells that express hOct4 , hNanog and hCripto ( Figure S2C ) , while SSEA4-negative cells were not reprogrammed . The observation that only a proportion of heterokaryons are successfully reprogrammed , as judged by hOct4 DNA demethylation and SSEA4 expression , might partly explain why the levels of transcripts encoding pluripotency factors are lower in reprogrammed cultures than established hES cell lines . To ask whether the reprogramming of human B cells by mouse ES cells resets multi-lineage potential , hB x mES cultures were treated with retinoic acid ( RA ) 6–8 days after cell fusion in order to induce differentiation ( Figure 3 ) . Prior to RA treatment most cells in hB x mES colonies showed alkaline phosphatase ( AP ) activity ( Figure 3A ) , and expressed human AP transcripts ( not shown ) . Hybrid colonies also expressed several pluripotency-associated markers , including hNanog protein ( detected using a human Nanog-specific antibody ) and the human embryonic-specific antigens SSEA4 , TRA-1-60 and TRA-1-81 [27] ( Figure S3 ) . Following treatment with RA , AP activity and expression of hOct4 , hNanog and hRex1 was reduced ( Figure 3C ) , while morphological heterogeneity within colonies increased . RA treatment induced the expression of genes associated with extra-embryonic ( hCdx2 , hHand1 and hGata6 ) , endoderm ( hSox7 , hHnf4 and hCollagenIVαI ) , mesoderm ( hMixl1 , hEbf and hMyoD ) and ectoderm ( hNestin , Figure 3B ) differentiation in hB x mES , but not in control hB cells ( Figure 3C , blue and black lines respectively ) . Differentiation also resulted in increased DNA methylation of the hOct4 promoter ( Figure 3D ) to levels similar with that seen in differentiated human cells ( Figure 1E ) . Taken together , these results show that reprogramming of human B cells by mouse ES cells resets gene expression and multi-lineage potential . Oct4 is part of the core regulatory circuitry in ES cells [35] and it is essential for pluripotency and self-renewal [36] . To assess the potential role of mouse-derived Oct4 as a dominant ‘trans’ acting factor within inter-species heterokaryons we generated ES cells expressing Flag-tagged mouse Oct4 protein ( Figure 4A ) and fused these with human B lymphocytes ( Figure 4B ) . Flag-tagged Oct4 ( derived from mouse ES cells ) was seen to accumulate within human nuclei 3 to 6 hours after cell fusion ( Figure 4B; complete kinetic experiment shown in Figure S4A ) . In addition , Oct4 protein was present in heterokaryon nuclei ( at 3 hours ) before transcription of hOct4 was initiated ( at 24 hours ) . Thus , translocation of the ES-derived Oct4 into human lymphocyte nuclei precedes reprogramming . Conversion of human fibroblasts to ES-like cells has been shown to require the activation of at least four factors including Oct4 , Sox2 and either Nanog plus Lin28 [9] or Klf4 plus c-Myc [10] . Recently it was shown that mouse ES cells lacking Sox2 , a factor thought to be vital for preventing extra-embryonic differentiation , can remain pluripotent provided with elevated Oct4 levels [37] . To investigate the relative importance of Oct4 and Sox2 in reprogramming , mouse ES cells that are inducible null ( Tet-off ) for mOct4 ( ZHBTc4 [36] ) or for mSox2 ( 2TS22C [37] ) were used as fusion partners with human B cells . These inducible null ES cell lines were constructed and characterised previously [36] , [37] and display a rapid ( within 24 hours ) and complete elimination of Oct4 or Sox2 gene/protein expression upon doxycycline ( +Dox ) treatment . In our hands , pre-treatment of ZHBTc4 cells with Dox for 6 and 12 hours , resulted in a progressive decrease in mOct4 gene expression ( Figure 4C ) , without significantly affecting the expression of other pluripotency-associated genes in these cells or the efficiency which they fuse with human B cells ( Figure S4B ) . Successful reprogramming , as judged by induction of several human genes ( Oct4 , Nanog , Cripto , Dnmt3b , Sox2 , Tle1 , Tert and Rex1 ) was however reduced ( +6 hours ) or eliminated ( +12 hours ) by pre-treatment of ZHBTc4 cells with Dox ( Figure 4D , a complete kinetic analysis is provided in Figure S4C ) . Likewise , knocking down mOct4 using short interference RNA ( siRNA ) in E14tg2a mES cells ( Figure S5A and Text S1 ) also abolished reprogramming activity ( Figure S5B ) . These results confirm that mOct4 expression is critically important for initiating successful reprogramming , in keeping with previous reports [7]–[10] , [38] . The extinction of human lymphocyte-specific genes was however not impaired by Oct4 removal ( Figure S4C ) , a result that may support previous findings that the activation and silencing of gene expression programs in heterokaryons are mechanistically distinct processes [13] . Eliminating mSox2 expression in the mouse ES cell ( Figure 5A , 2TS22C ) had , in contrast , a relatively mild effect on reprogramming efficiency ( Figure 5B , compare values at 0 , 12 and 24 hours of Dox treatment ) . Furthermore , reprogramming was fully restored in fusions using 2O1 cells , a Sox2-deficient mES cell line in which mOct4 expression is up-regulated [37] ( Figure 5A , B values shown in red and complete kinetics shown in Figure S6 ) . These data show that Oct4 , but not Sox2 , is critical for the dominant reprogramming activity of mouse ES cells . Interestingly , using 2O1 cells we observed the enhanced induction of hSox2 ( Figure 5B , red arrow ) , a result that suggests that mouse-derived Oct4 levels may be important for initiating hSox2 expression in somatic nuclei . To assess whether gene expression by the reprogrammed cell is stable ( self sustaining ) or requires the continuous supply of factors provided by the mouse ES cell , we generated hybrid cells between lymphocytes and ES cells in which Oct4 expression could be conditionally withdrawn ( ZHBTc4 , experimental outline depicted in Figure 6A ) . In these experiments fusions were performed between mouse lymphocytes carrying a silent , Oct4-driven GFP transgene ( GOF18ΔPE ) and mouse ZHBTc4 ES cells , to allow successfully reprogrammed hybrid cells to be identified on the basis of GFP re-expression by day 10 ( Figure 6A ) . Hybrid clones contained a rearranged IgH locus , consistent with their derivation from mouse B cells ( Figure S7A and Text S1 ) , displayed twice the DNA content of diploid cells ( 4n , Figure 6B ) and were karyotypically stable over the study period ( not shown ) . As anticipated , hybrid cells expressed ZHBTc4-derived Oct4βgeo transcripts and several pluripotency-associated genes , but did not express B cell markers such as CD19 , Pax5 and Ly108 ( Figure S7B ) . Two hybrid clones were selected for study ( hybrid 4 and 12 ) and were treated with Dox to selectively ablate expression of ZHBTc4-derived Oct4βgeo ( Figure 6C; Figure S7C shows the strategy used to selectively detect Oct4βgeo transgene expression ) . Withdrawal of ZHBTc4-derived Oct4 did not alter the expression of mNanog and mSox2 in reprogrammed cells ( Figure 6C ) , and did not precipitate differentiation towards trophectoderm or the up-regulation of mCdx2 and mHand1 expression [36] ( Figure 6D and Figure S7D ) ; events that are induced by the removal of Oct4 from the parental ZHBTc4 line ( Figure 6D right hand panel and Figure 6E ) . Thus , our data show that reprogramming of lymphocytes by mouse ES cells induces an epigenetically stable ( and heritable ) resetting of gene expression in the lymphocyte nucleus .
In this study we show that reprogramming human lymphocytes can be achieved using mouse ES cells as a cell fusion partner , a process that induces the re-expression of endogenous human genes normally associated with human blastocyst development and human ES cell lines . Successful interspecies reprogramming is initiated in heterokaryons prior to chromosome intermixing , and generates cells that express human FGF signalling pathway components and human ES-specific surface molecules such as SSEA4 , TRA-1-60 and TRA1-81 . We show that this reprogramming is critically dependent upon Oct4 , since Oct4 deletion abolishes the reprogramming capacity of mES cells . Conversion of human B cells into ES-like cells results in the re-modelling of the somatic genome with loss of DNA methylation at the hOct4 locus . Importantly , once reprogramming is initiated by factors produced by the dominant ( ES ) nucleus , we show that withdrawal of mOct4 does not compromise the phenotype of hybrid cells . This result implies that the reprogrammed state , once initiated , is both self-sustaining and heritable . One surprising aspect of the reprogramming data shown here is the rapidity of gene conversion and DNA demethylation that occurs within heterokaryons . As successful reprogramming is only achieved in a proportion of heterokaryons ( <13% ) , it is likely that partial DNA replication ( or repair ) is required for lymphocyte conversion . Previous studies have shown that reprogramming in experimental heterokaryons using adult cells from different lineages [13] , [39] , can be initiated before genome duplication and cell division . Here we show that conversion of unipotent lymphocytes towards multipotency is achieved in transient heterokaryons prior to cell division . Re-activation of human Oct4 and Nanog by human nuclei , has been shown to occur rapidly upon DNA de-methylation and Tpt1 activation induced by Xenopus oocytes [21] , [40] . The rapid re-activation of endogenous pluripotency-associated genes seen in inter-species heterokaryons is consistent with transgene re-activation studies that have reported Oct4gfp expression by MEFs [41] or NSCs [22] fused with mouse ES or EC cells . Collectively these results may have an impact for generating human ES-like cells . Proof that mouse ES cells can dominantly reset the multi-lineage potential in human somatic cells , together with evidence that this process begins prior to nuclear fusion , suggests that improved methods for removing mouse chromosomes from heterokaryons [42] may be applicable for generating human stem cell lines . Alternatively , using conditionally targeted mouse ES cells to dissect the roles of individual proteins thought to be critical for multipotent reprogramming , may provide a rationale for using distinct protein cocktails to directly re-set lineage potential . In the experiments presented here we have shown that reprogrammed human cells express a profile of transcripts , signalling molecules and surface antigens that are similar to those seen in human ES cells , and different from mouse ES cells . This suggests that an early human embryonic “program” of gene expression is initiated in human nuclei by trans-acting ( mouse ) factors . Differences between the expression profiles of mouse ES cells and human reprogrammed nuclei probably reflect discrepancies in cis-acting regions between the mouse and human genomes . In agreement with this idea , a study in which the entire hTert gene was introduced into mice , showed expression of the transgene was similar to endogenous hTert in humans , rather than mouse endogenous mTert [43] . It is interesting to speculate that some of the well-publicised differences between human and mouse ES cells may indeed reflect intrinsic species dissimilarities , rather than temporal differences in stem cells isolation [44] , [45] . We show that after fusion of human lymphocytes to mouse ES cells ( that are Lif and Bmp dependent ) , human ES-like cells are generated that express FGF signalling components ( and are not dependent of Lif/Bmp ) . Thus , our data suggest that differences between human and mouse ES cells may reflect distinct signalling and transcriptional networks , rather than necessarily when or where they were isolated during embryogenesis . We show here that Sox2 , in contrast to Oct4 , is not required to convert human lymphocytes into a multi-potent state . This observation contrasts with results obtained previously using iPS strategies to reprogram mouse and human fibroblasts [7]–[10] , [38] , mouse hepatocytes and stomach cells [46] and mouse B-lymphocytes [47] . Whether this is because of differences relating to the overexpression of transcription factor cocktails used in iPS , or that reprogramming occurs over an extended time period ( pluripotency-associated genes such as Oct4 , Nanog and Sox2 are reactivated after 2 weeks of transduction [48] , [49] ) , is not known . However , as Sox2 was recently shown to be dispensable for the activation of Oct–Sox enhancers in mouse ES cells [37] , it is also possible that additional Sox family members such as Sox4 , Sox11 and Sox15 , may have redundant functions with Sox2 in reprogramming . Interestingly , by enhancing Oct4 levels in Sox2-deficient ES cells ( ES-2O1 ) we show elevated expression of hSox2 by reprogrammed human B cells . Recent genome-wide studies have shown that Sox2 is a direct target of Oct4 in both human [35] and mouse [50] ES cells , a fact that could explain why hSox2 is efficiently reprogrammed using ES cells that overexpress mOct4 . In our hands , overexpression of exogenous Oct4 in lymphocytes did not induce pluripotent conversion ( Pereira & Terranova , unpublished results ) , a finding that argues that additional chromatin remodelling factors , perhaps including those known to interact with Oct4 [51] , [52] or associated with the process of DNA demethylation , may be critical for successful reprogramming . Collectively , our data show that interspecies heterokaryons can provide an interesting and complimentary approach to iPS , allowing the factors that are required to directly induce pluripotency to be defined individually and in combination .
EBV-transformed hB clones were maintained in RPMI supplemented with 10% foetal calf serum ( FCS ) , 2 mM L-glutamine and antibiotics ( 10 µg/ml Penicillin and Streptomycin ) . The Abelson transformed Oct4-GFP B-cell line was derived from the Oct4-GFP transgenic mice ( GOF18ΔPE ) [53] bone marrow , cloned and grown in RPMI supplemented with 20% FCS , non-essential amino acids , L-glutamine , 50 µM 2-mercaptoethanol , antibiotics and IL-7 ( 5 ng/ml; R&D systems , Minneapolis , MN ) . Mouse ES cells were grown and maintained undifferentiated either on irradiated SNL feeder layers ( E14Tg2a , Hprt−/− ES cells; CCE and E14 ) or directly on 0 . 1% gelatin-coated surfaces ( OS25 , ZHBTc4 and 2TS22C feeder-free ES cell lines ) . ES cells were grown in KO-DMEM medium plus 10% FCS , non-essential amino acids , L-glutamine , 2-mercaptoethanol , antibiotics and 1000 U/ml of leukaemia inhibitory factor ( ESGRO-LIF ) . Feeder-free ES cell lines were cultured in GMEM-BHK21 medium plus 10% FCS , non-essential amino acids , sodium pyruvate , sodium bicarbonate , 2 mM L-glutamine , 2-mercaptoethanol , antibiotics and 1000 U/ml of LIF . Doxycycline ( 1 µg/ml , Sigma ) or Retinoic acid ( 10−6 M , Sigma ) were added to the media when indicated . The Flag-mOct4 cell lines were derived by the overexpression of Flag-tagged mouse Oct4 in E14tg2a ES cells . Briefly , mouse Oct4 cDNA was cloned in the pDFLAG-cDNAIII vector ( Invitrogen ) . The cDNA , including two flag sequences at the 5′ end , was excised and sub-cloned into a suitable vector for expression in ES cells ( pCBA ) , with expression driven by the chicken β-actin promoter . The vector was then linearised and transfected by electroporation into mouse ES cells . G418 selection ( 400 µg/ml; Invitrogen ) was applied 48 hrs after and resistant clones were manually picked and screened by Western blot . Human ES cell lines H1 , H7 and H9 cells [27] were cultured in medium conditioned by mitotically inactivated MEFs supplemented with 8 ng/ml bFGF ( Peprotech , London , UK ) on matrigel-coated plates , as previously described [54] . Cells were routinely passaged at a 1∶3 dilution by treatment with 200 U/ml collagenase IV ( Invitrogen , Carlsbad , CA ) and mechanical dissociation . Heterokaryons were generated by fusing ES cells and B-lymphocytes using 50% polyethylene glycol , pH7 . 4 ( PEG 1500; Roche Diagnostics , Mannheim , Germany ) . Briefly , ES cells and hB-lymphocytes were respectively labelled with Vibrant 1 , 1′-dioctadecyl-3 , 3 , 3′ , 3′ tetramethylindodicarbocyanine ( DiD ) and 1 , 1′-dioctadecyl-3 , 3 , 3′ , 3′-tetramethylindocarbocyanine perchlorate ( DiI ) cell labelling solutions ( Molecular Probes , Eugene , OR ) . Cells were resuspended at 1×106 cells/ml in DMEM and labelled with 5 µl/ml of dye at 37°C , 15 min . ES and hB were then mixed in an appropriate ratio ( ES∶hB ratio 1∶1; ES∶Oct4-GFPB ratio 1∶5 ) , and were washed twice in PBS . The supernatant was completely removed and 1 ml of PEG ( 37°C ) was added to the pellet of cells over 60 sec and incubated at 37°C for 90 sec with constant stirring . Then , 4 ml of serum-free medium ( DMEM ) were carefully added over a period of 3 min , followed by 10 ml of DMEM and incubation at 37°C for 3 min . After centrifugation ( 1350 rpm , 5 min ) , the pellet was allowed to swell in complete medium for 3 min . Cell mixtures were then resuspended and cultured under conditions promoting the maintenance of undifferentiated mouse ES cells at 0 . 5×106 cells/cm2 . To eliminate unfused hB cells , Ouabain ( 10−5 M; Sigma ) was added to the medium 4 hours after cell fusion . When OS25 , ZHBTc4 and 2TS22C cell lines were used , proliferating ES cells were eliminated by the addition of 10−5 M Ara-C ( Cytosine β-D arabino furanoside; Sigma ) 4–6 hours after fusion and then removed after 16 hours . When E14tg2a ES cells or derivatives were used , HAT ( 20 µM hypoxanthine , 0 . 08 µM aminopterine and 3 . 2 µM thymidine; Sigma ) was added to the medium 24 hours after fusion . RNA extraction was performed using RNA-BEE reagent ( Tel-Test Inc . , Friendswood , TX ) and residual DNA was eliminated using the DNA-free kit ( Ambion , Austin , TX ) . 3 µg of total RNA was then reverse transcribed using Superscript First-Strand Synthesis system ( Qiagen ) with oligo ( dT ) 12-18 ( Invitrogen ) . cDNAs of interest were then quantified using real-time qPCR amplification . Real-time PCR analysis was carried out on a Opticon DNA engine using Opticon Monitor software ( MJ Research Inc . , Waltham , MA ) , running the following program: 95°C for 15 min , then 40 cycles of 94°C for 15 sec , 60°C for 30 sec , 72°C for 30 sec , followed by plate-read . PCR reactions included 2× Sybr-Green PCR Mastermix ( Qiagen ) , 300 nM primers and 2 µl of template in a 35 µl reaction volume . Each measurement was performed in triplicate and data normalised according to Gapdh expression . Primers were designed with Primer Express software ( Applied Biosystems ) and tested for the specific detection of human transcripts ( and not mouse ) . Standard curves were calculated on serial dilutions of positive control cDNA . Primer sequences used for this analysis are indicated in Table S1 . Bisulfite modification of DNA was carried out with the EZDNA methylation kit ( Zymogenetics Inc . , Orange , CA ) according to manufacturer's recommendations . PCR primers that recognise bisulfite-converted human DNA only are listed in Table S1 . Amplified products were cloned into pCR2 ( Invitrogen ) and ten clones were randomly picked and sequenced . For immunofluorescence and FACS analysis , the following antibodies and dilutions were used: mouse monoclonal anti-human Lamin A/C ( VP-L550; Vector Laboratories Inc . , Burlingame , CA ) at 1∶100 dilution; rabbit polyclonal anti-GFP ( A11122; Molecular Probes ) at 1∶200 dilution; mouse monoclonal anti-human SSEA4 ( MC-813-70; Developmental Hybridoma Studies Bank , Iowa City , IA ) at 1∶3 dilution; mouse monoclonal anti-human TRA-1-60 and TRA-1-81 ( MAB4360 and MAB4381; Chemicon International , Temecula , CA ) at 1∶12 and 1∶20 dilutions , respectively; rabbit polyclonal anti-human Nanog and Nestin ( ab21624 and ab28944; Abcam Ltd . , Cambridge , UK ) at 1∶100 dilution; mouse monoclonal anti-Flag ( F3165 , Sigma ) at 1∶1000 dilution . Secondary antibodies conjugated with fluorochromes were purchased from Molecular Probes and used at 1∶400 dilution . Immunofluorescence was performed as previously described [13] . Mouse and human nuclei were distinguished in the resulting heterokaryons by counterstaining with 4 , 6-diamidino-2-phenylindole ( DAPI ) and human Lamin A/C staining . Individual cells were delineated by F-actin staining ( Phalloidin; A12380 , Molecular Probes ) . For alkaline phosphatase assays , hybrid colonies 8 days after cell fusion were stained with alkaline phosphatase assay kit ( Sigma ) . All slides were analyzed on a Leica TCS SP5 confocal microscope and processed with Leica software and Adobe Photoshop . Images of live GFP fluorescent hybrid colonies and alkaline phosphatase staining were collected using a Leica DM IRE2 microscope running Metamorph software . For FACS analysis a FACScalibur ( BD Biosciences ) with CellQuest software was used . FACS purification was performed using a FACSAria cell sorter . Western blot analysis was performed as previously described [55] using a goat anti-Oct3/4 polyclonal antibody ( sc-8628; Santa Cruz Biotechnology Inc . , Santa Cruz , CA ) or a mouse anti-Flag monoclonal antibody . As a loading control , blots were incubated with anti-Lamin B polyclonal antibody ( sc-6216; Santa Cruz Biotechnology Inc . ) . Each lane contained 20 ìg total protein . | One of the most pressing objectives of medical research today is the development of approaches to restore the function of tissues damaged by accident or disease . An important goal for this work is the isolation of stem cell populations to replace missing or nonfunctioning cells . Because problems of immune rejection are likely to occur unless the recipient and donor stem cells are very closely matched , a desirable strategy is to convert differentiated cells ( such as white blood cells ) from patients into immature tailored stem cell populations . Here , we have experimentally fused human white blood cells and mouse embryonic stem cells and shown that this reprograms them to become stem-like . This kind of “differentiation reversal” is shown to be rapid and stable . It requires the stem cell–specific factor Oct4 , but does not require Sox2 . This approach allows the identification of factors that are required to reprogram human blood cells with the long-term perspective to eventually generate patient-specific stem cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology/stem",
"cells",
"cell",
"biology/developmental",
"molecular",
"mechanisms",
"developmental",
"biology/cell",
"differentiation",
"genetics",
"and",
"genomics/epigenetics",
"cell",
"biology/gene",
"expression"
] | 2008 | Heterokaryon-Based Reprogramming of Human B Lymphocytes for Pluripotency Requires Oct4 but Not Sox2 |
Remodelling the methylome is a hallmark of mammalian development and cell differentiation . However , current knowledge of DNA methylation dynamics in human tissue specification and organ development largely stems from the extrapolation of studies in vitro and animal models . Here , we report on the DNA methylation landscape using the 450k array of four human tissues ( amnion , muscle , adrenal and pancreas ) during the first and second trimester of gestation ( 9 , 18 and 22 weeks ) . We show that a tissue-specific signature , constituted by tissue-specific hypomethylated CpG sites , was already present at 9 weeks of gestation ( W9 ) . Furthermore , we report large-scale remodelling of DNA methylation from W9 to W22 . Gain of DNA methylation preferentially occurred near genes involved in general developmental processes , whereas loss of DNA methylation mapped to genes with tissue-specific functions . Dynamic DNA methylation was associated with enhancers , but not promoters . Comparison of our data with external fetal adrenal , brain and liver revealed striking similarities in the trajectory of DNA methylation during fetal development . The analysis of gene expression data indicated that dynamic DNA methylation was associated with the progressive repression of developmental programs and the activation of genes involved in tissue-specific processes . The DNA methylation landscape of human fetal development provides insight into regulatory elements that guide tissue specification and lead to organ functionality .
Methylation of CpG dinucleotides in the mammalian genome is a key epigenetic mark . Adult tissues have highly distinct genome-wide DNA methylation signatures consistent with the regulation of cell differentiation by epigenetic mechanisms [1–3] . Differences in DNA methylation between tissues have been shown to mark differences between germ layers [4] , preferentially at regions with low CpG content [2 , 5 , 6] , at enhancers [4] and alternative promoters [7 , 8] . Multiple studies have reported on the reprogramming of the human methylome during preimplantation embryo development [9–11] . In line with previous data on mice [12] , in humans DNA methylation is largely erased after conception , the paternal genome being actively and the maternal genome passively demethylated , to become remethylated with the implantation of the embryo [9 , 10 , 13 , 14] . However , systematic and detailed reports on DNA methylation dynamics during human fetal development remain scarce [15] , while such data is key to understand how epigenetic mechanisms drive tissue specification and organ functionality . Current views of fetal DNA methylation dynamics are largely extrapolated from studies on the differentiation of human and mouse cells in vitro [7 , 15–21] , and the comparison of differentiated tissues to human induced pluripotent stem cells and human embryonic stem cell lines [15] . An exception is fetal brain development in humans , for which recently reported in vivo data showed significant DNA methylation remodelling [15 , 22] . Recent developments in technology for interrogating genome-wide DNA methylation at single-nucleotide resolution [23] and detailed functional annotation of the human genome [24 , 25] provide an opportunity to chart DNA methylation during development and assign biological roles to the regions involved . Taking advantage of these developments , we report on DNA methylation dynamics during human fetal development of one extraembryonic tissue and three organs relevant for complex human diseases . This organ-specific catalogue of DNA methylation during development provides fundamental insights into processes guiding human development , but also into the biological function of non-coding regions , which are emerging as important from genome-wide association studies ( GWASs ) of complex diseases [26] . In addition , this catalogue may serve as a reference for studies on the role of epigenetic mechanisms in the association between an adverse prenatal environment and adulthood disease [27] since DNA methylation marks may have an heightened sensitivity for environmental perturbations during remodelling [28] .
To study DNA methylation dynamics in human fetal development , amnion , muscle , adrenal and pancreas samples of 11 fetuses were obtained at 9 , 18 and 22 weeks of gestation ( W9 , W18 and W22; S1A Fig ) . Genome-wide DNA methylation was investigated with the Illumina 450k array resulting in data on 452 , 490 CpG sites after quality control [29] ( S1B–S1E Fig ) . The study included three biological replicates per tissue and time point , except for W22 amnion ( n = 2 ) and W22 pancreas ( n = 2 ) ( S2A Fig ) . We first assessed differences in overall DNA methylation patterns between time points and tissues using hierarchical clustering based on Euclidean distance ( Fig 1A ) and multidimensional scaling ( MDS ) ( Figs 1B and S2B ) . DNA methylation patterns clearly differentiated the four tissue types studied ( Figs 1A , 1B and S2B ) . The amnion , representing an extraembryonic tissue , clustered separately from the three embryonic tissues ( Fig 1A ) . Within the embryonic cluster , all W9 tissues ( representative of the first trimester ) clustered together , whereas W18-W22 tissues ( representative of the second trimester ) were present towards the edges of the MDS plot ( Figs 1B and S2B ) . Despite the distinct differences in DNA methylation patterns from the first to second trimester , the total number of hypo- , intermediately and hypermethylated CpGs remained constant across time points and tissues ( including the extraembryonic amnion ) on both autosomes and , in females , the X chromosome ( S2C and S2D Fig ) . This suggests that the observed differences in the MDS plot were not driven by changes in average levels of DNA methylation , but rather due to tissue- and time-specific changes in DNA methylation . To validate our findings , we integrated our data with three previously published Illumina 450k datasets on 10 human fetal tissues [15 , 22 , 30] . Hierarchical clustering of all data together ( n = 117 ) confirmed the presence of distinct tissue- and time-specific DNA methylation patterns in fetal tissues ( S2E Fig ) . The characteristics and biological function of DNA methylation depend on the local CpG content and position relative to genes [31] . We mapped CpG sites ( CpGs ) to CpG islands ( CGIs as defined in the UCSC genome browser; 138 , 919 CpGs ) , their shores ( ±2 kb of CGIs; 103 , 453 CpGs ) and shelves ( ±2 kb of shores; 42 , 227 CpGs ) , and remaining CpG-poor non-CGI regions ( 157 , 560 CpGs ) , and to regions relative to gene locations including distal promoters ( -10 kb–-1 . 5 kb; 21 , 101 CpGs ) , proximal promoters ( -1 . 5 kb–+0 . 5 kb; 171 , 077 CpGs ) , gene bodies ( +0 . 5 kb–3’ untranslated region ( UTR ) ; 175 , 062 CpGs ) , downstream regions ( 3’ UTR–+5 kb; 8 , 563 CpGs ) and remaining intergenic regions ( 66 , 356 CpGs; Fig 1C ) . CpGs were commonly hypomethylated in CGIs , intermediately methylated in shores and hypermethylated in both shelves and non-CGI regions ( Fig 1C ) . These patterns differed by genic position , e . g . CGI methylation was lowest in proximal promoters and highest in gene bodies . Annotation-specific methylation differences were found between W9 and W22 , as CpGs in CGIs and shores tended to increase ( e . g . gene body CGIs and distal promoter shores ) , whereas CpGs in non-CGI regions decreased in methylation ( e . g . non-CGI promoters ) . For a subset of annotations , the amnion showed a slightly different DNA methylation patterns than for embryonic tissues , e . g . for gene body CGIs ( Fig 1C ) . Taken together , our data imply that DNA methylation is highly dynamic during fetal development without affecting the average level of DNA methylation . It has been shown that each adult tissue is defined by tissue-specific DNA hypomethylation [15 , 32 , 33] . Since the four fetal tissues analysed showed a clear DNA methylation signature that corresponded to separated clusters ( Fig 1A and 1B ) , we investigated whether combinations of tissue-specific DNA hypomethylated CpGs were present irrespective of its developmental stage . To do this , we identified CpGs that were relatively hypomethylated ( defined as a DNA methylation difference of > 20% ) in each tissue compared to all others throughout the three time points of fetal development investigated . The analysis showed indeed that , independently of the developmental age , each tissue showed a cluster of tissue-specific hypomethylated CpGs ( Fig 2A ) . The early lineage segregation of the amnion was further confirmed by the comparatively large number of CpGs ( 3 , 536 CpGs ) that were exclusively hypomethylated across amniotic samples . In contrast , the embryonic tissues contained much fewer tissue-specific hypomethylated CpGs ( muscle 756 CpGs; adrenal 140 CpGs; pancreas 220 CpGs ) reflecting their common origin of the epiblast , that gives rise to all embryonic tissues . Genes mapping ( i . e . the nearest gene locus ) to the specific hypomethylated CpGs per tissue regardless of the time point ( amnion 2372 , muscle 548 , adrenal 120 , pancreas 175 ) were enriched for biological processes that included GO terms characteristic of amnion , muscle and pancreas development and function ( S1 Table ) . When annotated to genic and CGI-related location , it became evident that tissue-specific hypomethylation was enriched for non-CGI regions ( P < 0 . 0001 ) and highly depleted at CGIs , in particular when mapping to proximal promoters ( P < 0 . 0001; Figs 2B and S3A ) . To gain further insight in the biological role of genomic regions displaying tissue-specific hypomethylation , we used chromatin state segmentations for fetal muscle , fetal adrenal , amnion and adult pancreatic islets generated by the Epigenomics Roadmap [25] . Tissue-specific hypomethylation was strongly enriched at enhancers ( P < 0 . 001; Figs 2C and S3B ) . The functional relevance of those tissue-specific hypomethylated CpGs was further validated by comparison to additional fetal samples [15] and adult somatic tissues [6] from available external datasets ( S3C Fig ) . Intriguingly , a high degree of similarity between tissues sharing the same origin , even into adulthood , was observed ( S3C Fig ) . We next investigated whether tissue-specific hypomethylated CpGs clustered into Hypomethylated Regions ( tHRs , defined as 3 consecutive hypomethylated CpGs within 1kb of each other ) [6] . This was the case for amnion , muscle and pancreas ( Tables 1 and S2 ) . tHRs comprise robust development-independent epigenetic markers as exemplified by the mapping of pancreatic tHRs to proximal promoters of the nearest genes ACY3 , HNF1A , and HNF4A ( Table 1 and S3D Fig ) , genes with a key role in pancreas development [34 , 35] . Muscle tHRs ( S2 Table ) mapped to distal elements of transcription factors involved in muscle development ( NFATC1 ) and somitogenesis ( UNCX ) [36 , 37] ( S3D Fig and Tables 1 and S2 ) . Importantly , we could confirm the tHRs identified with the relatively sparse Illumina 450k array with fetal and adult muscle whole-genome bisulfite sequencing ( WGBS ) data ( S3E Fig ) [25] . These data indicate that it is feasible to use combinations of tHRs as tissue-specific and development-independent barcodes . We provide evidence for large-scale DNA methylation dynamics between W9 and W22 ( DNA methylation difference > 20% ) that affected 11 . 5% of evaluated CpGs ( 52 , 134/452 , 490 ) . Approximately equal numbers of CpGs showed a gain of methylation ( GOM ) ( 26 , 555 CpGs; amnion 5 , 988; muscle 7 , 631; adrenal 13 , 997; pancreas 8 , 620 ) and a loss of methylation ( LOM ) ( 25 , 579 CpGs; amnion 10 , 811; muscle 11 , 925; adrenal 4 , 476; pancreas 3 , 286 ) ( Fig 3A ) . DNA methylation remodelling occurred predominantly between W9 and W18 and not between W18 and adulthood ( Figs 3B and S4A ) . Intriguingly , the integration and re-analysis of external DNA methylation data of fetal adrenal , brain and liver [6 , 15 , 22 , 30 , 38] revealed a striking confirmation of DNA methylation dynamics during fetal development . Furthermore , for all embryonic tissues , the DNA methylation levels at W22 were similar to those found in the adult counterpart ( Figs 3B and S4A ) , suggesting that the extent of changes after W22 are limited for these CpGs . GOM CpGs did not show tissue-specific patterns ( Fig 3A ) and , in line with this observation , often corresponded to genes involved in generic developmental and cellular processes , including embryonic morphogenesis and regulation of transcription ( S3 Table ) . In contrast , the LOM CpGs were highly tissue-specific ( Fig 3A ) and mapped to genes involved in tissue-specific processes that matched the organ in which the LOM CpGs were identified ( S3 Table ) . CpGs that lost methylation in the amnion mapped , amongst others , to genes that were associated with the regulation of apoptosis and cytoskeleton organization; in the muscle to genes associated with cytoskeleton organization and muscle system processes; in the adrenal to genes associated with regulation of macromolecule metabolism ( S3 Table ) . In the pancreas no significant enrichments were found . GOM and LOM CpGs differed in their genomic annotation . While GOM CpGs were generally enriched in CGIs and CGI-shores , LOM CpGs were enriched for CGI-shelves and non-CGI regions ( Figs 3C and S4B ) . LOM- and GOM-specific enrichments were also observed for Epigenomics Roadmap chromatin state segmentations . LOM CpGs were strongly enriched for ( genic ) enhancers and transcribed regions , whereas GOM CpGs were enriched for bivalent and repressed regions and only modestly at enhancers ( Figs 3D and S4C ) . The results underscore the relevance of DNA methylation in enhancer activity , in addition to the well-studied relationship between DNA methylation and promoter activity [39] . We previously reported on transcriptional data of amnion ( n = 7 ) , muscle ( n = 6 ) , adrenal ( n = 3 ) and pancreas ( n = 7 ) at W9 , W18 and W22 [40] and used this data to test the hypothesis that GOM is associated with the epigenetic downregulation of developmental programs and LOM with upregulation of tissue-specific processes . Genes associated with GOM and involved in embryonic morphogenesis ( a process enriched for GOM in all tissues ) showed a decrease in transcriptional activity from W9 to W22 in all tissues ( amnion , muscle , pancreas: P < 0 . 05; adrenal P = 0 . 87; Fig 4A ) . In contrast , genes involved in tissue-specific processes found to be enriched for LOM ( S3 Table ) increased in transcription from W9 to W22 ( P < 0 . 05; Figs 4B and S4D ) . Altogether , these findings emphasize that DNA methylation dynamics during human fetal development is associated with the availability to transcription of both general embryonic programs ( shutting those down for transcription ) as well as tissue-specific developmental programs ( making those available for transcription ) . From the dynamically methylated CpGs , we identified 2 , 229 development-related differentially methylated regions ( dDMRs , defined as 3 consecutive differentially methylated CpGs within 1kb of each other ) undergoing GOM ( amnion 185; muscle 530; adrenal 1 , 065; pancreas 449 ) and 1 , 017 undergoing LOM ( amnion 388; muscle 482; adrenal 136; pancreas 61; S4 Table ) . After mapping the dDMRs to the nearest gene locus we observed that the percentage of common genes in the embryonic tissues showing LOM dDMRs was 1 . 3% , whereas those showing GOM was 10 . 2% ( Fig 5A ) . LOM dDMRs were associated with genes involved in tissue-specific functions such as MYH3 in muscle ( muscle contractile protein ) , MC2R ( adrenocorticotropic hormone receptor ) in adrenal and PFKFB3 ( involved in insulin secretion ) in the pancreas ( Table 2 and S5A Fig ) . As an example of tissue specificity encountered in the dDMRs showing a loss of DNA methylation , we zoomed in on the MYLK2 locus , a muscle-specific gene [41] . The methylation of the MYLK2 promoter and first exon in the muscle decreased during development , but increased ( or remained constant ) in the other organs studied ( Fig 5B ) . Interestingly , GOM dDMRs were associated with ( tissue-specific ) developmental genes , such as PAX3 in muscle , a key gene in myogenesis [42] , and NKX6 . 1 in pancreas , an important gene in beta-cell development [43] ( Table 2 and S5A Fig ) , but also near well-known developmental genes including the HOXB ( Fig 5C ) and other HOX clusters ( S6A Fig ) that play a key role in embryonic patterning and morphogenesis [44] . However , when comparing all identified dDMRs to previously identified adult ( tissue-specific ) tDMRs using the 450k array [6] , about 50% of the GOM dDMRs were not identified as tDMRs in adult tissues ( S5B Fig ) , while 32% and 38% , of the LOM dDMRs in muscle and pancreas , respectively , were unique for those fetal tissues ( adult data on adrenal was absent ) . The persistence into adulthood of subsets of GOM and LOM dDMRs was confirmed using WGBS data for adult muscle [25] ( S5C Fig ) . These results suggest that the study of DNA methylation dynamics in fetal development will identify regions that are remodelled during development and are missed when studying adult tissues only . We further explored the potential biological validity of the 1 , 012 muscle dDMRs using ENCODE data [24] on human skeletal muscle myoblasts ( HSMMs ) and their differentiated derivatives , human skeletal muscle myotubes ( HSMMtubes ) . In HSMMs and HSMMtubes , DNAse I hypersensitive sites ( DHSs ) , which mark genomic regions of open chromatin associated with transcriptional activity , were abundant at LOM dDMRs , particularly in CpG-poor regions ( CGI-shelves and non-CGI regions , Fig 5D left ) . DHSs were depleted at GOM dDMRs in CpG-rich regions ( CGIs and CGI-shores , Fig 5D left ) . Consistent with an increased transcriptional activity , LOM dDMRs were also enriched in myotubes and myoblasts ( ENCODE [24] ) for histone H3 lysine 4 methylation ( ( H3K4me1 , -me2 , -me3 ) [25 , 45] , and acetylation of histone H3 at lysine 9 and 27 ( H3K9ac , H3K27ac ) , all marks associated with active regulatory regions [25 , 45] ( Fig 5D right ) . In contrast , these active histone modifications were depleted for GOM dDMRs in CpG-rich regions ( Fig 5D right ) . LOM dDMRs were depleted of H3K9me3 ( marking inactive DNA ) , H3K27me3 ( marking Polycomb-repressed regions ) and H3K36me3 in HSMMtubes but not in their precursor cells HSMMs ( Fig 5D right ) . A final indication for the functional relevance of the muscle dDMRs was that 124 out of 482 LOM dDMRs significantly overlapped ( P < 0 . 0001 ) with binding sites of the muscle-specific transcription factor MYOD in HSMMs ( 188/482 in HSMMtubes ) , whereas only 7 out of the 530 GOM dDMRs mapped to MYOD binding sites[46] ( 8/530 in HSMMtubes; S5D Fig ) .
Here , we show that human tissues already exhibit a specific DNA methylation signature as early as W9 of fetal development . In addition , the DNA methylation landscape is subjected to considerable changes from the first to second trimester of gestation as the developing organs gain complexity and functionality . Our study highlights that dynamic DNA methylation is not only an integral part of early preimplantation embryo development and implantation [9–11] , but continues to be a key feature of epigenetic remodelling during human fetal development . While global changes in levels of DNA methylation characterize development until implantation ( Fig 6 ) , these are not observed during fetal development . Instead , distinct LOM occurs near tissue-specific genes and GOM occurs near developmental genes in a largely tissue-independent fashion ( Fig 6 ) . Our direct assessment of DNA methylation dynamics suggests that a larger proportion of the methylome is remodelled during development than previously thought [3 , 6 , 47] . Interestingly , the functional relevance of identified dynamic regions was further exemplified by the changes in expression of their nearest genes . While the nearest genes of regions gaining DNA methylation associated with embryonic morphogenesis showed loss of expression , the nearest genes of regions losing DNA methylation showed increased expression over time . In agreement with our observations , LOM of hematopoietic-specific genes has been observed during human hematopoietic differentiation in vitro [48] and have been linked to transcriptional changes in human T-cell development [49] . Moreover , several mouse and human in vitro studies demonstrated that the methylation of developmental genes increases [19 , 50] and tissue-specific functional genes lose methylation [5 , 51] during stem and progenitor cell differentiation . Lastly , DNA demethylating agents , such as 5-azacytidine , have been shown to promote stem cell differentiation and maturation of skeletal myotubes in mice [52 , 53] . Further experimental studies are required to evaluate the mechanistic role of DNA methylation in development . Although our study reveals general principles of DNA methylation dynamics during human fetal development , it should be noted that a limited number of tissues and individuals was investigated; and that we used a genome-wide method interrogating a relatively small proportion of all CpGs in the human genome . Expansion to more tissues and the application of whole-methylome technologies will lead to a more comprehensive catalogue of regulatory regions . However , by extensive inclusion of external fetal and adult 450k array datasets , we have consolidated our findings . Moreover , the use of external available WGBS data confirmed the results obtained by the 450k array data . Since we studied organ biopsies similar to previous studies investigating biopsies of human adult tissues [3 , 15] , the methylation profiles we report reflect the average of multiple cell types . The cellular complexity of the organs investigated led to an underestimation of the actual DNA methylation dynamics in individual cell types . This is exemplified by the detection of a considerably larger number of CpGs displaying dynamic methylation in muscle which has an exclusive mesodermal origin in comparison with adrenal and pancreas which are composed by cells originating from two different germ layers ( adrenal: mesoderm and ectoderm ( neural crest ) ; pancreas: endoderm and mesoderm ) . However , it is unlikely that the methylation dynamics observed is an epiphenomenon of this cellular complexity instead of being driven by cell differentiation and maturation . This is obvious for genes associated with GOM that appears to be shared across organs to repress general developmental programs during development . In contrast , genes associated with LOM displayed tissue-specific patterns . Their intricate involvement in organ-specific functions was emphasized by tight linkage to biological processes and chromatin states relevant to the organs investigated . Moreover , between W9 and W22 , the organs analysed are mainly composed of progenitor cells; perfusion by blood and lymphatic vasculature , and innervation by neural crest cell derivatives still plays a minor role as compared with adult organs ( S1A Fig ) . In the future , single-cell methodology [54 , 55] will enable comparing single-cell DNA methylomes of the various adult cell types to their fetal progenitor counterparts . Studies of DNA methylation landscapes of human fetal development may serve as reference in the development of ( organoid ) differentiation models [56] and , moreover , shed light on potential mechanisms underlying genetic associations and studies in the field of epigenetic epidemiology [57] focussing on the prenatal environment .
The Medical Ethical Committee of the Leiden University Medical Center approved this study ( P08 . 087 ) . Informed consent was obtained on the basis of the Declaration of Helsinki ( World Medical Association ) . Human fetal tissues ( amnion , skeletal muscle , adrenal glands , pancreas ) at gestational age W9 , W18 , W22 ( S2A Fig ) were collected from elective abortion material ( vacuum aspiration ) without medical indication . In this study , “weeks of gestation” was used as determined by the last menstrual period ( LMP ) . After collection , the material was washed with 0 . 9% NaCl ( Fresenius Kabi , France ) and the identified organs were immediately snap-frozen using dry ice and stored at -80°C until further processing . Histology was performed as previously described [58] . The images were taken with an Olympus AX70 microscope ( Olympus , Japan ) provided with a XC50 digital colour camera ( Olympus , Japan ) . Tissues were homogenized with a pestle and lysed overnight at 56°C with proteinase K ( 600 mAU/ml , Qiagen , Germany ) in ATL buffer ( Qiagen , Germany ) . After lysis , residual RNA in the samples was degraded using RNase A ( 10 mg/μl , Invitrogen , USA ) . Subsequently , genomic DNA ( gDNA ) was extracted on the basis of phenol/chloroform . Briefly , lysates were transferred to Phase Lock Heavy Gel 2ml Eppendorf tubes ( 5PRIME , Germany ) and 700 μl of 25:24:1 Phenol/Chloroform/Isoamyl alcohol was added and spun down for 5 minutes . The aqueous phase was transferred to a Phase Lock tube and the latter step was repeated . The aqueous phase was transferred to a new Phase Lock tube and 700 μl 24:1 Chloroform/Isoamyl alcohol was added and spun down for 5 minutes . The aqueous phase was transferred to a Phase Lock tube and the latter step was repeated . The aqueous phase was transferred to a new 2 μl Eppendorf tube ( Eppendorf AG , Germany ) , 70 μl 3M sodium acetate ( Ambion , USA ) and 1400 μl ice cold 100% ethanol were added . gDNA was precipitated over night at -20°C . Eppendorf tubes were spun down at 4°C for 15 minutes and washed twice with 70% ethanol . After the pellet was dry , gDNA was solubilized in AE buffer ( Qiagen , Germany ) and stored at 4°C . DNA concentration was determined using the Qubit dsDNA BR Assay Kit on a Qubit 2 . 0 Fluorometer ( Invitrogen , USA ) . gDNA was bisulfite converted using the EZ-96 DNA methylation kit ( Zymo Research , Orange County , USA ) with an average input of 600 ng gDNA . Following bisulfite conversion , DNA methylation data was generated using Illumina HumanMethylation450 BeadChip according to the manufacturer’s protocol . All analyses were performed using R statistics , version 3 . 0 . 1 . The 65 polymorphic SNP probes featured on the 450k array were used to exclude potential sample mix ups . Data was imported in R using minfi [59] and processed and normalized using a custom pipeline: Arrays were removed if they had a low median intensity , a high background signal or with incomplete bisulfite conversion , but none were excluded ( S1B , S1C and S1D Fig ) . The CpGs on chromosome X and Y were used to confirm sex ( S1C Fig ) . Next , probes with a low bead count ( < 3 ) , high detection P-value ( > 0 . 01 ) , and with a low success rate ( < 95% ) , and ambiguously mapped probes [60] were removed . After probe filtering , all arrays contained > 95% of the original number of probes . Background correction and colour correction were applied and the data was quantile normalized ( lumi [61] ) . To adjust for the type I/II bias BMIQ was applied [62] . In our analyses , CpG sites in the sex chromosomes ( Y and X ) were excluded . High correlation was found between samples from the same time point and tissue but also between time points of the same tissue ( S1D Fig ) . To exclude chromosomal abnormalities , we calculated the copy number aberration based on the signal intensities using the method published by Feber et al . [63] as implemented in the R package ChAMP [64] . From these results , no abnormalities were found in the samples used ( S1E Fig ) . Multidimensional scaling and clustering was performed based on Euclidean distance . For the DNA methylation over time within features , a genic annotation was combined with a CGI-centric annotation as presented before [6] to determine the median methylation per combined feature . P were calculated using quantile regression based on the median ( R package quantreg [65] ) . Figures were made using the R-packages ggplot2 [66] and GenomeGraphs [67] . Tissue-specific hypomethylation: CpGs with a standard deviation ≥ 0 . 1 within the tissue of interest were discarded . Relative tissue-specific hypomethylation was defined as hypomethylation of the tissue of interest compared to the other tissues , with a difference of ≥ 0 . 20 in beta value . CpGs sites were selected if a difference was consistent in each of the time points . Dynamic methylation: CpGs with a high standard deviation ≥ 0 . 10 within time point/tissues indicative of an instable estimate of DNA methylation were discarded from this analysis . Gain and loss of methylation was defined a gain/loss of ≥ 0 . 20 between W9 and W22 and W18 in between the two time points ( W18 was allowed to be 0 . 05 lower/higher in beta value than W9/W22 respectively ) . The CpGs with a gain or a loss of methylation were used for the combined genic/CGI-centric annotation and expressed as an odds ratio . Chromatin state segmentation data [25] were downloaded from the Epigenomics Roadmap Project for fetal muscle ( W15 female ) , fetal adrenal ( W13 male ) , amnion ( W16 male ) and pancreatic islets ( adult ) and enrichment of dynamically methylated CpGs was calculated . DMRs: In both , the relative hypomethylated CpGs ( tHRs ) and the CpGs with a gain or a loss of methylation ( dDMRs ) , DMRs were called using an algorithm described before [6] . Briefly , DMRs ( tHRs and dDMRs ) were defined by three consecutive CpGs that matched a criterion ( that is , relative hypomethylation or gain/loss of methylation ) with a maximum of 1 kb between CpGs and with at highest three CpGs that did not match the criterion . Gene ontology: Tissue-specific hypomethylated and dynamically methylated CpGs were mapped to their nearest gene ( that is to the nearest TSS or TES of a gene ) and tested for enrichment of gene ontology terms using DAVID [68] . For the tissue-specific hypomethylation we used a P cut-off of 0 . 05 on the raw P as the number of CpGs was relatively low . For the dynamically methylated CpGs a FDR cut-off of 0 . 05 was set as cut-off for enrichment in GO terms . A background set was used containing nearest genes of all CpGs covered on the array . Gene expression data: Transcriptional data of the four tissues at W9 , W18 , W22 ( amnion: n = 2 , 3 , 2 , muscle: n = 2 , 2 , 2; adrenal: n = 1 , 1 , 1; pancreas: n = 3 , 2 , 2 ) were used . The counts per million ( CPM ) expression levels were calculated using the R package edgeR 3 . 2 . 4 [69 , 70] . For the plots , the arithmetic mean of the biological replicates was used and the median of all genes plotted . To access enrichment of up- and downregulation , a probability test was used . MYOD , DNAse I , histone marks and WGBS: Overlaps between MYOD binding peaks and muscle dDMRs were calculated . To test for significance , we calculated an empirical distribution by performing 20 , 000 permutations with 482 ( gain of methylation ) and 530 ( loss of methylation ) DMR-like regions each and determined the overlap with the MYOD binding sites . DMR-like regions were defined as regions with equal characteristics as dDMRs identified: an inter-CpG distance smaller than 1 kb and an average length of five CpGs per DMR-like region ( n~8 x 104 regions ) . The two-sided P was determined using the empirical distribution . DNAse I and histone mark data of human skeletal muscle myoblasts ( HSMMs ) and human skeletal muscle myotubes ( HSMMtubes ) were downloaded from the ENCODE website [24] . DNAse I hypersensitivity was expressed as the count of DNAse-seq tags . The enrichment of histone marks was expressed as the log2 of the ChIP/input . The total number of reads within the myotubes was different from the total number of reads in the myoblast data and , therefore , the data was normalized . dDMRs were classified as island ( CGIs and their shores ) or non-island dDMRs , and DNAse-seq tags and histone marks around dDMRs were mapped up to 5 kb up- and downstream . CpG sites of WGBS data were mapped to hypomethylated and dynamic regions and their 5kb flanking regions . Using a smooth spline , the methylation around the regions was smoothed for the adult and fetal data . Methylation data has been deposited in the NCBI’s Gene Expression Omnibus [71] under accession number GSE56515 . External datasets that have been used in this manuscript include: fetal and adult DNA methylation data of various tissues from Nazor et al . ( Gene Expression Omnibus ( GEO ) accession number: GSE31848 ) [15] , fetal brain DNA methylation data from Spiers et al . ( GEO accession number: GSE58885 ) [22] , fetal liver DNA methylation data from Bonder et al . ( GEO accession number: GSE61279 ) [30] , adult DNA methylation data of various tissues from Slieker et al . ( GEO accession number: GSE48472 ) [6] , fetal Deep SAGE expression data of the four tissues studied here from Roost et al . ( GEO accession number: GSE66302 ) [40] , adult DNA methylation brain data from Pidsley et al . ( GEO accession number: GSE61431 ) [38] , WGBS data of fetal and adult muscle generated by the Epigenomics Roadmap consortium ( GEO accession numbers: GSM1172596 and GSM1010986 ) , MYOD binding peaks from MacQuarrie et al . ( GEO accession numbers: GSM1218849 and GSM1218850 ) [46] . | Methylation of DNA is a key epigenetic mark . Adult tissues have highly distinct genome-wide DNA methylation signatures . How these signatures arise during human fetal development is largely unknown . Here , we studied DNA methylation profiles of four tissues ( amnion , muscle , adrenal , pancreas ) during first and second trimester of human fetal development . Already in the first trimester , a tissue-specific signature was found in each of the tissues . However , during the first and second trimester , a substantial number of genomic regions were found to gain and lose DNA methylation . Genomic regions that gained methylation were associated with the shut-down of developmental processes , while genomic regions that lose methylation were associated with the activation of tissue-specific functions . These findings on the DNA methylation landscape of human fetal development are important as they provide insight into regulatory elements that guide tissue specification and lead to organ functionality . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | DNA Methylation Landscapes of Human Fetal Development |
Injury to nerve axons induces diverse responses in neuronal cell bodies , some of which are influenced by the distance from the site of injury . This suggests that neurons have the capacity to estimate the distance of the injury site from their cell body . Recent work has shown that the molecular motor dynein transports importin-mediated retrograde signaling complexes from axonal lesion sites to cell bodies , raising the question whether dynein-based mechanisms enable axonal distance estimations in injured neurons ? We used computer simulations to examine mechanisms that may provide nerve cells with dynein-dependent distance assessment capabilities . A multiple-signals model was postulated based on the time delay between the arrival of two or more signals produced at the site of injury–a rapid signal carried by action potentials or similar mechanisms and slower signals carried by dynein . The time delay between the arrivals of these two types of signals should reflect the distance traversed , and simulations of this model show that it can indeed provide a basis for distance measurements in the context of nerve injuries . The analyses indicate that the suggested mechanism can allow nerve cells to discriminate between distances differing by 10% or more of their total axon length , and suggest that dynein-based retrograde signaling in neurons can be utilized for this purpose over different scales of nerves and organisms . Moreover , such a mechanism might also function in synapse to nucleus signaling in uninjured neurons . This could potentially allow a neuron to dynamically sense the relative lengths of its processes on an ongoing basis , enabling appropriate metabolic output from cell body to processes .
Neurons extend extremely long axonal processes that can exceed the diameter of the cell body by 4–5 orders of magnitude . This poses a unique challenge for intra-cellular signaling , since nerve cells require efficient transport mechanisms to move macromolecules and metabolites from the cell body to neurite terminals and back over distance . This communication problem becomes especially acute in the context of nerve injury , when the axon needs to provide the cell body with accurate and timely information regarding the site and extent of axonal damage [1] . Cell body responses to axonal injury are diverse , ranging from functional repair to cell death , and depend on both the intrinsic regeneration capacity of the neuron and responses to the local environment [2]–[4] . The distance of the lesion site from the cell body is one of the factors determining neuronal responses to injury . For some populations of neurons , a more proximal axotomy leads to greater regenerative response by the cell body ( [5]–[7] and references cited therein ) . Lesion distance was also shown to influence specific molecular responses to injury , including activation of cell body kinases [8] and up-regulation of growth-associated genes [5] , [9]–[12] . Interestingly , the precise effect of lesion distance on neuronal response may differ in diverse neuronal populations . For example , an optic nerve lesion study reported that the number of regenerating retinal ganglion cells is inversely correlated with distance of the lesion from the optic disc [13] . In long neurons from two species of fish , lesions close to the cell body induce death , while beyond a certain lesion distance neurons regenerate [14] , [15] . Moreover , the lag time for initiation of regeneration in these neurons is tightly correlated with lesion distance [14] , [15] . Taken together , these findings demonstrate that neurons from different functional classes and species have the capacity to differentiate between lesion sites at different locations in their axons . Early workers in the field proposed a number of hypotheses to explain disparate cell body responses to differently located axonal lesions [16] , [17] . Diffusion mediated signaling is not likely to function efficiently over the requisite distances [18] , and other mechanisms like signaling waves [19] or spatial gradients of protein abundance [20] have not been demonstrated to occur over axonal distances . On the other hand , two long distance signaling mechanisms have been characterized in nerve injury paradigms- a rapid electrophysiological signal of short duration [21] and a second slower wave of signals transported on molecular motors [1] , [22] . Motor-driven signaling has emerged as a versatile mechanism for long distance communication along nerve axons [23] , [24] , and in this study we have used computer simulations to examine the possibility that it can provide lesion distance information in injured neurons . The analyses support feasibility of a multiple signals model , wherein distance information is inferred from the time delay between the arrival of an electrophysiological fast signal and slow signals carried by the molecular motor dynein . The simulations indicate that this mechanism can enable nerve cells to distinguish between distances of 10% or more of their total axon length .
Inferring the distance traveled by a given signal can rely on two types of mechanisms , either quantifying chemical gradients over distance , or measuring the time delay between initiation of the signal and its arrival in the detection region . Although chemical gradients play central roles in biological systems , diffusion-based gradients cannot be established over axonal distances within a biologically relevant time frame after injury [25] . Thus , we examined mainly the second possibility , namely that a time delay between the initiation of a chemical signal in an injured axon and its arrival at the cell body can be interpreted by the cell as representing the distance traveled by the signal . In order for such a mechanism to work , it requires two reference points: an early time point representing the initiation of the signal , and a later time point representing the arrival of the signal . The latter requires a detection system that responds to the arrival of an amount of signal defined by a specified threshold , while for the former requirement , we hereby suggest two models that can in principle define signal initiation: Both models are based on measuring the arrival of a sufficient amount of the slow signal , defined as a fraction of 500 in silico particles moving in a Matlab-defined simulation environment ( see methods ) . Since we do not have any data regarding the signal concentration required for initiating a response , our models explore a series of thresholds , defined as fractions of arriving signal from the total signal generated at the injury site . These sensitivity thresholds range between 1% and 90% of the injury signal ( i . e . , a 1% detector will respond to 1% of the originally generated signal , whereas accumulation of 90% of the signal is required for response of a 90% detector ) . Retrograde transport along the microtubule cytoskeleton in nerve axons is almost entirely dynein-based , thus the basic assumption in our models is that the slower signals are carried retrogradely as part of a dynein based complex . Dynein velocities have been measured in diverse systems from isolated molecules in vitro to in intact cells , and by different methods including direct imaging or end-point accumulation , leading to reports of a range of velocities from ∼0 . 5 µm/sec to ∼5 µm/sec [23] , [27] . Our models require inputs of velocity distributions ( rather than average velocities ) , and we therefore extracted velocity distributions from two experimental data-sets , one based on in vitro analyses of movement of individual dynein–dynactin–GFP complexes [28] , and another that utilized cellular imaging of the retrograde transport of a GFP-labeled endosome marker in embryonic motor neurons [29] . For both data sets , a curve fit procedure was applied ( see methods ) resulting in a distribution function . Based on these distribution functions , random velocity values were assigned to migrating particles simulating dynein-trafficked retrograde signals . Figure 2 depicts the experimental data and the fitted distribution functions for both data sets . Unless otherwise specified , all simulations utilized the distribution function derived from the data of Ref . [28] . As depicted in Figure 3A , our initial model system performs a comparative measurement . Two injuries are performed in two distinct cells in-silico . In one cell the injury is introduced in a proximal location , and in the other cell a distal injury is performed . In response to each of the two injuries , a slow signal emanates from the injury sites , propagating retrogradely towards the cell body . The system then measures the time delay between the fast and slow signals from the proximal location ( Δt1 ) and the time delay between the fast and slow signals from the distal location ( Δt2 ) . The difference between Δt2 and Δt1 reflects the system's ability to distinguish between the two locations: the larger this difference , the better the system in terms of distance measurement . The simulations explore the influence of two parameters on system performance: the distance between the two injuries ( hereafter referred to as injury displacement ) , and the total distance between the distal injury and the detector ( L , or total distance ) . Figure 3B depicts a schematic representation of two cases , one in which the displacement is small and the total distance is relatively large , and one in which the displacement is relatively large compared to the total distance . The intuitive prediction is that a biological system will find it more difficult to distinguish between the two injury sites in the former case rather than in the latter . In order to assess consistency of model performance , we repeated each such in-silico experiment 100 times . In each such repetition , the same detector sensitivity , same total distance from cell body , and same injury-displacement distance were used . Differences between repetitions emerge solely from random assignment of dynein velocities to the slow signal particles . Figure 3C depicts two sets of such 100 repeats for the two-signals model ( the same procedure was applied to the two-detectors model , data not shown ) . In the first case ( Figure 3C , left panel ) , the parameters that were chosen were: L ( total distance ) = 60 cm , Disp ( injury displacement ) = 0 . 5 cm , and the detector-sensitivity threshold was set to 30% . Each vertical bar ( dark blue ) represents a single repetition of the simulation . The value obtained for each repetition represents the time difference Δt2−Δt1 in minutes . The negative bars observed for approximately one tenth of the repeats indicate that for these specific simulations the system infers mistakenly that the distal injury site is closer to the cell body than the proximal injury site . In another ∼10% of the repeats , the measured Δt2−Δt1 time difference in signal arrival is less than an hour . Since the different molecular events involved in both generating the retrograde injury signals at the site of injury and interpreting them at the cell body may take about an hour [30]–[32] , such a time difference in signal arrival might be below the resolving power of an injured neuron ( i . e . , even though the signal from the proximal injury site traveled up to an hour less than the signal from the distal site , the accompanying events of signal production and/or processing may exceed this time difference , thus making it biologically irrelevant ) . Moreover , despite the fact that these are 100 repeats of the same injury and displacement distances , reproducibility of the measurement is clearly very poor . Thus , at least for this 0 . 5 cm displacement distance that is two orders of magnitude smaller than the 60 cm total injury distance , the initial model cannot discriminate between locations of the two injury sites . In the second case ( Figure 3C , right panel ) , we set L to 5 cm and Disp to 3 . 5 cm , using the same sensitivity threshold of 30% . In this case , the system provided a consistent set of measurements , all ranging around 16–18 hours . Figure 3C depicts two extreme examples of model performance for two distinct combinations of total distance/injury displacement . In order to conduct a systematic exploration of model performance , we extended this analysis to cover a wide range of distance-displacement combinations . For each distance-displacement combination , we performed 100 simulation repeats as described above . From each such set of 100 simulations we discarded the worst 5% , and then chose the minimal Δt2−Δt1 time-difference value out of the remaining 95% of the repetitions ( Figure S1 , red circle ) . Note that in the examples of Figure 3C this minimum is a negative value for the left panel , while in the case of the right panel the minimum value is approximately 1000 minutes . We then used the collection of minima points to plot a 3D graph in which the X and Y axes represent injury displacement and total distance , respectively , and the Z axis represents the minimal Δt2−Δt1 time difference value for each X–Y combination ( Figure S1 , lower panel ) . Such graphs can be used to answer two basic questions regarding the models- first , can a given model distinguish between two distinct injury locations . This is determined by setting a cutoff for system failure due to either mistaken identification of the distal injury site as being closer than the proximal ( resulting in a negative Z axis value ) , or a time delay that is too small to enable a differential biological response . Since differential biological responses to injury typically require transcription and translation , for purposes of the modeling the system cutoff was defined as a time delay of at least 60 minutes . The second issue addressed by the 3D plots is whether a given model is consistent , i . e . will it provide a similar assessment for the same injury displacement , regardless of its distance from the cell body ? This is reflected in the smoothness of the graph . In an ideal system , the time difference in the arrival of a signal that travels a distance x and a signal that travels a distance x+Δx should remain constant , regardless of the value of x . Thus for an ‘ideal’ 3D graph ( Figure S2 ) , straight lines along the X axis indicate consistency ( i . e . , for a given value of injury displacement , the time difference ( Z ) should remain the same at all total distance values ) . In order to assess the smoothness of a 3D graph plotted from the simulations , we use a root mean square deviation ( RMSD ) measurement . Given two sets of n points v and w , the RMSD is defined as follows:When calculating the RMSD for a model-generated graph compared to an ideally smooth graph , the lower the obtained RMSD value , the closer the graph to the ideal , hence the effects of changing parameters and models can be inferred from their comparative RMSD values . Systematic exploration of the two-signals model showed that although it can function over part of the total distance/injury displacement combinations , the system failed over a significant portion of the distances range tested ( Figure 4A , Figure S3 ) . Furthermore , for a given injury displacement , the time-delay measurements did not show consistency over increasing total-distance values . For example , the ability of the system to detect an injury displacement of 8 cm decays with distance along the axon , and is essentially lost at total distances of 70–80 cm and above . RMSD values for a wide range of detector sensitivity thresholds indicate that the system performed better at sensitivity settings of up to 30% , and worsened significantly in the range from 40% to 80% ( Figure S3 ) . Performance of the two-detectors model was much poorer , and in the best case the system detected injury location differences for only approximately one third of total distance/injury displacement values ( Figure 4B , Figure S4 ) . Unfortunately , the RMSD measurement seems to be uninformative for comparing different permutations of the two-detectors model . Rather than reflecting model performance , RMSD values reflect the ‘gap’ between the sensitive detector and the insensitive detector . The larger the difference between the thresholds of the two detectors , the larger the time difference between the distal and proximal locations . Thus , two 3D graphs that are similar in terms of smoothness , but differ in their Z values ( time differences ) will yield different RMSD values ( Figure S4 ) . Since model performance in two signals or two detectors mode was not satisfactory , we modified the two-signals model to include several slow signals rather than a single slow signal , and assume that an effective response is triggered when a subset of these signals arrives at the cell body ( Figure 5 ) . From a biological point of view , this may reflect a situation in which there are several dynein-carried signals . We further assume that as far as detector-sensitivity is concerned , there is no significant difference between the signals ( i . e . , in terms of our model they utilize similar detection systems ) . The rationale behind this modification is that in a noisy system , multiple measurements are expected to be more accurate than a single measurement . In its original configuration , in order for a distal injury to be identified by the system as a proximal one , it was sufficient that a small fraction of the slow signal particles emanating from the distal site would randomly acquire higher velocities than the signal particles originating from the proximal point . In order for a similar phenomenon to occur in the multiple signals system , the distal point needs to randomly “win” not only once , but in several slow-signal velocity acquisitions . Figure 6 compares the performance of a system with six slow signals , of which any three will initiate a response , versus performance of the previously described system with a single slow signal . A significant improvement is observed in consistency ( graph smoothness ) , together with a marked increase in the total distance and injury displacement ranges for which the system attains a successful outcome ( Figure 6A and Figure S5 ) . RMSD values are also significantly improved ( Figure S5 ) . We considered examining a similar extension of the two-detectors model to multiple detectors . However , whereas extending the two-signals model to multiple signals did not require any new ( and unjustified ) assumptions regarding system parameters , a similar extension of the two-detectors model requires overly speculative assumptions . Consider , for example , a system with three kinds of detectors with sensitivity thresholds s1 , s2 , and s3 , where s1<s2<s3 ( i . e . , s1 is the most sensitive detector ) . The limiting determinant of system performance will have to be the time delay between activation of s1 and s3 – having s2 as an intermediate detector will not influence the result , unless one assumes preferential effects of such intermediate detectors . In the absence of any data , such speculative configurations may be completely detached from biological reality . Nonetheless , we did try to modify some quantitative features of the slow signal , in order to check whether the poor performance of the two-detectors model results from the specific biological datasets that were used in this work . We applied the following modifications to the model: ( i ) using a uniform distribution of velocities instead of the data-based Gaussian distributions , ( ii ) using velocities 1–3 orders of magnitude faster than the data-based velocities , and ( iii ) using wider and narrower velocity distributions ( obtained by modifying the parameters of the curve-fit functions described in the Methods section below ) . None of these modifications yielded any significant improvement in model performance ( data not shown ) . It therefore seems that the the two ( or multiple ) signals model is qualitatively superior to the two-detectors model , and the difference in model performances cannot be attributed to a quirk of specific model configuration . As noted above , we used two sets of dynein velocity measurements for our modeling work: a data-set from Ross et al . [28] , representing velocities of isolated dynein-dynactin complexes in vitro , and a data set from Deinhardt et al . [29] , based on tracking of GFP-labeled tetanus toxin in live motor neurons . The average dynein velocity measured by Deinhardt et al . was higher than the average dynein velocity measured by Ross et al . – 1 . 3 µm/sec and 0 . 45 µm/sec , respectively , and the velocity distributions of Deinhardt et al . spanned a broader range ( Figure 2 ) . As a consequence , time delays between the arrival of signals from distal and proximal locations in simulations based on the Deinhardt et al . data were smaller than in simulations based on the Ross et al . data , and simulations based on the Deinhardt et al . data were more susceptible to noise ( Figure 7 ) . Thus , in a system configuration integrating five out of ten signals ( a model configuration based on multiple slow signals – see also Figure 5 and accompanying text above ) , simulations based on the Ross et al . data yield satisfactory results over a broader combination of distances and injury displacements than simulations based on Deinhardt et al . 's data ( Figure 7 ) . Nonetheless , increasing the number of signals and detector sensitivities for the Deinhardt et al . data show the same trends for improvement as demonstrated for simulations based on Ross et al . ( data not shown ) . Thus , it is reasonable to assume that optimal results can be obtained also from relatively noisy motor behavior given a sufficient number of signals and appropriate detector sensitivity .
The distance between the site of injury and the cell body seems to have a significant effect on a neuron's ability to recover from mechanical injury [14]–[17] , [33] , [34] . Furthermore , there are both qualitative and quantitative aspects to this distance effect . In specific neuron types , once distance between cell body and site of injury drops below a certain lower threshold , no regeneration occurs , whereas above this threshold the probability of regeneration increases continuously with the increase in distance between the cell body and site of injury [14] , [15] . In other neuronal populations , a more proximal axotomy leads to greater regenerative response by the cell body [5]–[7] . Despite the clear biological significance of injury distance in neural tissues , the mechanism by which distance from the site of injury is measured is unknown , and the degree of precision required from such a measurement is not clear . In this work , we aimed at providing a theoretical framework for examining how intracellular distance measurement might be accomplished at the cellular level within a neuron . Computer simulations based on existing biological data were used to examine these concepts , and to assess their plausibility . Nonetheless , we are fully aware that the results and conclusions presented in this paper were derived from models that are abstractions of the real biological system , although we tried to keep speculations regarding the the mechanisms driving the behavior of these models to the bare minimum . We should also note some of the limitations of our approach , thus for example dynein velocity might be influenced by the type of cargo [35] . Although this was not factored into our models , the analyses show that the differences between the two velocity distributions used for model simulations do not affect key qualitative behaviors of the system ( Fig . 7 ) . Another issue not explicitly modeled is processivity of the dynein motor , namely the propensity of the motor to stall , or to move over limited distances in the opposite direction [28] , [36] . In the above described simulations , signaling molecules were assigned a given velocity , and they continued moving retrogradely with that velocity throughout the entire simulation . We carried out initial tests of the effects of motor pausing behaviors by running simulations at which in each time step 30% of the particles were randomly selected to remain in the same position until the next time step ( Figure S6 ) . This modification did not seem to have any significant effect on model performance . We further examined the effect of switching velocities in the model by re-assigning velocities to 10% of the molecules once per 100 time steps ( a typical simulation is of the order of 104 time steps ) . As can be seen in Figure S7 , this modification improved the performance of the system in terms of failure percentage . This can easily be understood by considering that if a given signaling molecule undergoes velocity switches for sufficient time , eventually the velocity of each molecule will converge to the average velocity of the entire population , decreasing noise in the system . Thus , our main findings without considering the possibility of velocity switching may actually reflect a worst-case scenario . Despite the above caveats , the modeling shows that in principle a set of dynein-mediated signals can provide intracellular distance information in an injured neuron . Furthermore , we did not have to add any “external players” to or impose speculative mechanisms on the model . Both the fast electrical signal and the slow chemical signal have been characterized in the context of nerve cell injury [21] . Moreover , such a mechanism might also function in synapse to nucleus signaling in uninjured neurons if a neurotransmitter or other synaptic stimulation elicits electrical ( fast ) signals concomitantly with dynein-based ( slow ) signals . Such a scenario has actually been reported for the neurotrophin BDNF , which elicits both rapid electrophysiological signals [37] and dynein-transported signaling endosomes [38] . NMDA receptor signaling provides another example , transmitting both acute electrophysiological signals [39] and activating macromolecule transport by importins and dynein [40] , [41] . If such signaling systems are indeed used to sense synapse to nucleus distance , this would allow autonomic length measurements of neuronal processes on an ongoing basis , which in turn could guide metabolic output from neuronal cell bodies to processes . The existence of cellular mechanisms that detect time delays between signaling events has been shown to exist in diverse biological systems ( e . g . [42] , [43] ) . Even the expansion of the model to multiple slow signals reflects the existence of multiple signaling complexes which are retrogradely transported by dynein [1] , [22] , [24] . The proposed model can fit a large range of nerve lengths , covering a diversity of organism sizes . Finally , the models allow two firm conclusions that might be testable experimentally in the future; first , that use of multiple and partly redundant signaling entities provides a more robust distance assessment mechanism measurement than a single signal , and second that distance detection resolution is proportional to neurite length ( Figure 8 ) . It will be intriguing to follow experimental testing of these ideas in the future .
In order to produce a velocity distribution function , a data fit procedure was applied to the two experimental datasets used in this work . Both datasets were obtained from analyses of the movement of individual molecular complexes – either in vitro [28] or in live neurons [29] . In both cases , the authors reported their results as the relative occurrence of given ranges of velocities ( e . g . , 5% of the observations were at velocity ranges of 0–0 . 2 microns/sec ) over a non-exhaustive number of molecular complexes ( 148 discrete complexes in Reference [28] , 126 in Reference [29] ) . Thus , the reported velocity sets are not an ideal representation of velocity distribution , but rather an experimentally limited sampling . The curve fit procedure allowed us to compute a continuous function which could then be used to randomly assign velocities to the signaling molecules in each simulation round . For this purpose , we used a built-in Matlab script ( fminsearch ) based on the Nelder-Mead method [44] , [45] . Since our model does not acount for zero velocities , we introduced a slight modification to the Gaussian function , thus requiring the velocity distribution function to intersect with ( 0 , 0 ) . The curve fit function that was used was of the form:Thus , for the velocity X = 0 , the function yields zero occurrences . Goodness-of-fit was assessed by calculating the root mean square deviation ( RMSD ) between the observed data points and the values predicted by the calculated function:where the values are given in terms of percentage-of-occurrence of given dynein velocities ( see Fig . 2 ) . This measurement provides an estimate for the average distance between a given data point and the calculated curve . For the Ross et al . data set , the following results were obtained:RMSD = 3 . 1% For the Deinhardt et al . data the following results were obtained:RMSD = 0 . 61% Molecular transport complexes are represented as moving particles . Each such particle has a location in space , and it can move according to its velocity . This approach also allows extension of the model in the future to include additional molecular properties and experimental data . In our model , a signal is composed of 500 moving particles . In order for a signal to achieve its effect , a minimal fraction of the signal should arrive at the detector , this is presented as detector sensitivity in % in the results section . The influence of various sensitivity thresholds was examined during simulations . All simulation scripts were written in MATLAB , and simulation executions were performed on the Wiccopt cluster ( hosted by The Weizmann Institute's computing center ) to allow parallel executions of simulations which varied in initial parameter settings . The Cluster's nodes consist of machines with: 2 quadcore xeon CPU's , 1 quadcore xeon CPU , 2 dualcore AMD opteron , and 1 dualcore AMD opteron . | Neurons have extremely long axonal processes that can reach lengths of up to 1 meter in human peripheral nerves . The neuronal cell body response to nerve injury is dependent on signals carried by molecular motors from the lesion site in the axon . The distance between the injury site and the cell body influences the type of response , suggesting that neurons must be able to estimate the distance of an axonal injury site , although how they do this is unknown . We have used a computational approach to model intracellular distance measurement after nerve injury . The models show the feasibility of a mechanism based on a rapid , near instantaneous , signal carried by action potentials in the nerve , followed by multiple slower signals carried on molecular motors . Such a mechanism can enable a neuron to discriminate between distances as close as 10% of total axon length . The model provides insights on retrograde injury signaling in neurons , including the biological relevance of the mechanism over different scales of nerves and organisms . Moreover , if similar mechanisms function in synapse to nucleus signaling in uninjured neurons , this could enable estimation of relative process lengths , thus guiding metabolic output from cell bodies to axons . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/signaling",
"networks",
"neuroscience/neuronal",
"signaling",
"mechanisms",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration",
"cell",
"biology/cell",
"signaling"
] | 2009 | Can Molecular Motors Drive Distance Measurements in Injured Neurons? |
Memory CD8 T cells confer increased protection to immune hosts upon secondary viral , bacterial , and parasitic infections . The level of protection provided depends on the numbers , quality ( functional ability ) , and location of memory CD8 T cells present at the time of infection . While primary memory CD8 T cells can be maintained for the life of the host , the full extent of phenotypic and functional changes that occur over time after initial antigen encounter remains poorly characterized . Here we show that critical properties of circulating primary memory CD8 T cells , including location , phenotype , cytokine production , maintenance , secondary proliferation , secondary memory generation potential , and mitochondrial function change with time after infection . Interestingly , phenotypic and functional alterations in the memory population are not due solely to shifts in the ratio of effector ( CD62Llo ) and central memory ( CD62Lhi ) cells , but also occur within defined CD62Lhi memory CD8 T cell subsets . CD62Lhi memory cells retain the ability to efficiently produce cytokines with time after infection . However , while it is was not formally tested whether changes in CD62Lhi memory CD8 T cells over time occur in a cell intrinsic manner or are due to selective death and/or survival , the gene expression profiles of CD62Lhi memory CD8 T cells change , phenotypic heterogeneity decreases , and mitochondrial function and proliferative capacity in either a lymphopenic environment or in response to antigen re-encounter increase with time . Importantly , and in accordance with their enhanced proliferative and metabolic capabilities , protection provided against chronic LCMV clone-13 infection increases over time for both circulating memory CD8 T cell populations and for CD62Lhi memory cells . Taken together , the data in this study reveal that memory CD8 T cells continue to change with time after infection and suggest that the outcome of vaccination strategies designed to elicit protective memory CD8 T cells using single or prime-boost immunizations depends upon the timing between antigen encounters .
Memory CD8 T cells provide immune hosts with enhanced protection from pathogenic infection due to an increased precursor frequency of antigen ( Ag ) -specific cells , widespread localization to both lymphoid and non-lymphoid tissues , and ability to rapidly execute effector functions such as cytokine production and cytolysis compared to naïve CD8 T cells [1–3] . Protection provided by memory CD8 T cells is dependent upon the number , quality ( functional abilities ) , and location of memory CD8 T cells available at the time of infection . Importantly , the quality and location of memory CD8 T cells best suited to combat diverse infections is dependent upon the tropism of the invading pathogen . Memory CD8 T cells consist of a heterogeneous population of cells [4] that were initially categorized into central memory ( Tcm ) and effector memory ( Tem ) subsets based on CCR7 and CD62L expression , and that differ in anatomical location and functionality [5 , 6] . Recently , an additional subset of memory CD8 T cells has been described that reside in non-lymphoid tissues and that have been called tissue-resident memory ( Trm ) cells [7] . While the relative protection provided by circulating Tcm and Tem cells differs depending on the nature of infection [6 , 8–10] , both are better suited to provide protection against systemic infection than Trm cells that provide enhanced protection against infection that occurs within peripheral tissues [11–15] . Several studies have suggested that Trm cells may be long-lived in the skin following VacV or HSV infection and in mucosal surfaces following intramuscular immunization with adenovirus vectors [12 , 15 , 16] . However , other studies examining Trm generated following influenza have suggested that Trm cell numbers wane following infection [17] . Therefore , longevity of Trm cells likely depends on the infection/vaccination model and the tissue of memory residence . However , circulating memory CD8 T cells persist for great lengths of time following immunization or systemic viral infection . For example , lymphocytic choriomeningitis virus ( LCMV ) -specific memory CD8 T cells are maintained at stable numbers in the spleen for the life of the laboratory mouse [18] , and detectable numbers of memory CD8 T cells can be found in human PBL 20–75 years after natural exposure to , or vaccination against yellow fever virus , measles virus , and smallpox [19–23] . However , several studies have indicated that some properties of circulating memory CD8 T cells change with time after infection . For example , expression of CD62L and CD27 ( markers of central memory cells ) increases , indicating that the subset composition of the memory population changes with time after infection . In addition , functions such as cytokine production , proliferation , and memory generation following Ag re-encounter , increase with time [24–28] . The full extent of phenotypic and functional alterations that occur within the memory CD8 T cell population with time after infection , however , remains poorly characterized . It is unclear if alterations are due solely to differences in subset composition of memory CD8 T cell populations , or to changes within defined memory subsets . These are important questions to address , as the level of protection provided against systemic infections may change with time following initial infection and/or vaccination . While most vaccines are intended to elicit protection against seasonal illnesses or pathogens that will be encountered in the relatively near future , infection may not occur for long periods of time following the original vaccination . Therefore , changes in memory CD8 T cell function between vaccination and the time of infection may impact the protection provided by memory CD8 T cells . Furthermore , evidence has suggested that the number of memory CD8 T cells required to provide protection against some pathogens may be quite high [29–31] . Currently , the best method for eliciting large numbers of memory CD8 T cells involves prime-boost strategies in which a series of vaccinations are administered allowing for a period of time between boosts [32] . Recently , it was reported that all human subjects receiving five injections of cryopreserved radiation attenuated Plasmodium falciparum sporozoites were protected upon controlled infection with malaria , while not all subjects receiving four injections were protected [31] . A number of potential reasons for the increased protection provided by the five-dose regimen were proposed including a longer interval of time between administration of the fourth and fifth boost . Although not tested in their work , the authors argued that increasing the interval of time between boosts could lead to the establishment of greater numbers of memory CD8 T cells and increased protection compared to immunization strategies using a shorter time interval between boosts . Thus , functional changes in the properties of memory CD8 T cells occurring in the time between boosts may directly impact the protection achieved through prime-boost vaccination strategies . For these reasons , an understanding of how memory CD8 T cell quality changes with time after infection and/or vaccination is needed . In this study we examined how the properties of circulating memory CD8 T cells change with time following an acute systemic infection with LCMV . We demonstrate that memory CD8 T cell quality changes with time after Ag-encounter in a manner not solely due to shifts in memory subset composition . Importantly , our data suggests that alterations in memory CD8 T cell function that occur with time after Ag-encounter could impact their ability to provide protection against diverse pathogens , and that the generation of memory CD8 T cells through prime boost protocols may depend on the timing between boosts .
Heterogeneous populations of Ag-specific CD8 T cells can be analyzed on the level of the population ( every CD8 T cell in the host ) , the subset level ( CD8 T cells expressing a marker or combination of phenotypic markers ) , or the level of single cells . Tcm and Tem subsets differ in anatomical location and functionality [5 , 6] . Thus , differences in function between memory populations could be due to alterations in subset composition that occur with time after primary antigen recognition . Additionally , individual cells within the population and within subsets can differ in phenotype and function from one another . However , because the level of protection is determined by the quality of all memory CD8 T cells present at the time of re-infection , we first examined how circulating memory CD8 T cells change with time after infection when analyzed on the population level . We adoptively transferred low numbers of naïve Thy1 . 1 or Thy1 . 1/1 . 2 transgenic ( Tg ) P14 CD8 T cells specific for the glycoprotein ( GP ) 33-41 epitope derived from LCMV into Thy1 . 2 C57BL/6 recipients and infected recipients with LCMV 24 hours ( h ) later . We then analyzed memory P14 cells on the population level ( i . e . all memory cells present in the examined organs ) 30–45 days ( earlyM ) or 8+ months ( lateM ) later . Although similar numbers were found in the spleen ( Fig 1A ) , earlyM cells were found in greater proportions in the lungs while lateM cells were found in greater proportions in the inguinal lymph nodes following perfusion of tissues ( Fig 1B ) . Surface marker profiles also differed with time , with expression of CD127 , CD62L , CD27 , and CD122 increasing and expression of KLRG1 decreasing with time after infection ( Fig 1C ) . Similar patterns of surface marker expression were also seen for endogenous earlyM and lateM GP33 and GP276 –specific CD8 T cell responses ( S1A Fig ) . To determine if cytokine production or degranulation changes with time , intracellular cytokine staining ( ICS ) was performed on earlyM and lateM cells that were mixed and incubated with GP33-41 peptide for 5 h . The percentages of IFN-γ and TNF-α producing P14 cells and degranulation as measured by surface CD107a expression did not change with time . However , the percentage of memory CD8 T cells able to produce IL-2 increased with time ( Fig 1D ) . Consistent with these data , the percentage of cells capable of polyfunctional cytokine production ( IFN-γ , TNF-α , and IL-2 ) increased with time after infection ( Fig 1E ) . Finally , to determine if maintenance of memory CD8 T cells changes with time , basal proliferation of earlyM and lateM cells was examined with bromodeoxyuridine ( BrdU ) incorporation during an eight-day period . Interestingly , a higher percentage ( p<0 . 01 ) of lateM cells incorporated BrdU during the eight-day interval ( Fig 1F ) , indicating that the rate of basal proliferation increases in memory CD8 T cells with time after infection . Basal proliferation is partially dependent upon IL-15 signaling [33–35] , and increased expression of IL-15Rβ ( CD122 ) in lateM cells ( Fig 1C ) suggested that sensitivity to IL-15 could change with time . To test this , Thy disparate earlyM and lateM cells were labeled with carboxyfluorescein diacetate succinamidyl ester ( CFSE ) , mixed and incubated with increasing concentrations of IL-15 , and dilution of CFSE was determined after 3 d . lateM cells proliferated to a greater extent in response to exogenous IL-15 and were more sensitive to lower levels of IL-15 ( Fig 1G ) , indicating that sensitivity to IL-15 increases over time . Expression of CD122 was also greater in lateM compared to earlyM endogenous GP33 and GP276 tetramer positive cells ( S1A Fig ) , suggesting that endogenous lateM cells are also more responsive to IL-15 compared to earlyM cells . These data show that while memory CD8 T cells analyzed on the population level persist at stable numbers in the spleen after infection , time changes their location , surface marker expression , Ag-driven cytokine production , and ability to respond to homeostatic clues in the environment . Primary memory CD8 T cells robustly proliferate and generate secondary effector and memory populations after Ag re-encounter [27 , 36] . Proliferation by primary memory CD8 T cells was reported to increase with time after initial Ag exposure following secondary infection with Sendai virus or Listeria monocytogenes [26 , 27] . To verify that time-dependent changes in Ag-driven proliferation and secondary memory generation of primary memory CD8 T cells analyzed on the population level are not dependent upon the type of infection or Ag-specificity , we set up adoptive co-transfer experiments . Thy disparate earlyM and lateM P14 cells were mixed ( Fig 2A ) and 2x104 of each were transferred into naïve C57BL/6 recipients followed by infection with various pathogens expressing GP33 . This experimental setup ensured that secondary responses generated from earlyM and lateM cells were subject to the same in vivo environmental conditions throughout the response . Secondary responses generated from earlyM and lateM P14 cells after LCMV Armstrong ( LCMV ) , LM , and Vaccinia Virus ( VacV ) infection were tracked longitudinally in peripheral blood ( PBL ) , and a significantly greater percentage of secondary effector cells was found for responses generated from lateM compared to earlyM P14 cells after each infection ( Fig 2B and 2C ) . Additionally , when progeny were examined at a secondary memory time point , we found a 5–7 fold increase in the percentage of secondary memory cells generated from lateM compared to earlyM P14 cells after each infection ( Fig 2C ) . Furthermore , a greater percentage of secondary effector and memory cells was found for responses generated from lateM compared to earlyM P14 cells in multiple organs following LCMV infection ( Fig 2D ) . These results indicate that reduced progeny generated from earlyM cells was not due to restriction of responses by primarily CD62L- earlyM cells to non-lymphoid tissues . Taken together , these data indicate that Ag-driven proliferation and memory generation potential of primary memory CD8 T cells analyzed on the population level increases with time after infection irrespective of antigen specificity or type of infection . Historically , circulating memory CD8 T cells have been characterized as CD62Llo Tem or CD62Lhi Tcm cells based on expression of CD62L and localization in peripheral tissues and lymphoid organs [5 , 6] . Functional differences between Tem and Tcm subsets also have been demonstrated , with Tcm having a greater capacity to produce IL-2 , increased proliferative potential , and the ability to provide increased protection following some , but not all infections [6 , 8] . Therefore , the functional changes observed in the memory CD8 T cell population with time could be due solely to changes in subset composition . Alternatively , the properties of defined subsets of memory CD8 T cells could change with time . To test the hypothesis that changes in memory CD8 T cell properties with time after infection are not due solely to shifts in subset composition , we began to examine the properties of earlyM and lateM CD62Lhi cells . We chose to focus on CD62Lhi memory subsets for two reasons . First , with time after acute infection memory CD8 T cells convert to a population that is primarily CD62Lhi [24 , 25 , 27 , 28] ( S2A Fig ) . Second , with time the CD62Llo population becomes enriched for T death intermediate memory cells ( TDIMs ) that arise from homeostatic division , are non-functional , and are destined to die [37] . To further document this , we examined IFN-γ production by CD62Lhi and CD62Llo earlyM and lateM P14 cells following stimulation with cognate Ag . The capacity of CD62Lhi earlyM and lateM cells to produce IFN-γ was similar ( S2B Fig ) . In contrast , a higher percentage of CD62Llo lateM compared to earlyM cells were unable to produce IFN-γ ( S2B Fig ) , consistent with an enrichment of TDIMs . At present , there is no reliable phenotypic marker that can be used to distinguish TDIMs from normally functioning memory CD8 T cells within the CD62Llo population . Thus , by focusing our analysis on CD62Lhi memory CD8 T cells , we were able to examine how a well-characterized subset of memory CD8 T cells changes with time after infection while avoiding complications in analyzing CD62Llo populations . Surface marker expression has been used to characterize memory CD8 T cells with different functional abilities . We rationalized that if CD62Lhi earlyM and lateM cells displayed changes in phenotype , they also were likely to display other differences as well . Therefore , we began to explore to what extent the phenotype of CD62Lhi memory populations was influenced by time by examining the expression of CD27 , CD127 , and CD122 , markers highly expressed by Tcm cells , on CD62Lhi earlyM and lateM P14 cells ( Fig 3A ) . The percentage of CD62Lhi cells expressing CD27 , CD127 , and CD122 was greater in lateM compared to earlyM P14 cells in the peripheral blood ( PBL ) and spleen indicating that the phenotype of CD62Lhi memory CD8 T cells continues to change with time . This pattern was also seen for expression of CD122 and KLRG1 on gated CD62Lhi endogenous GP33 and GP276 earlyM and lateM cells ( S1B Fig ) . The percentage of cells expressing CD27 , CD127 , and CD122 among CD62Lhi earlyM P14 cells also differed between organs . Percentages of cells expressing CD27 and CD127 in the lungs were lower compared to the PBL , spleen , and inguinal lymph nodes , while percentages of cells expressing CD122 were lower in the inguinal lymph nodes compared to the PBL , spleen , and lungs . In contrast , percentages of CD62Lhi lateM cells expressing CD27 , CD127 , and CD122 were uniformly high and were similar regardless of anatomical location except for a reduction in the percentage of cells expressing CD122 in the inguinal lymph nodes . This suggested that phenotypic heterogeneity within the CD62Lhi memory CD8 T cell subset decreases with time after infection . To further test this , we co-stained for expression of CD27 , CD127 , CD122 , CD11b , and KLRG1 on CD62Lhi earlyM and lateM P14 cells . This strategy allowed us to differentiate 32 different subpopulations of CD62Lhi memory P14 cells , and the percentage of each subpopulation of total CD62Lhi earlyM and lateM P14 cells is shown in S3A Fig . To examine heterogeneity within CD62Lhi earlyM and lateM cells , the number of subpopulations ( out of 32 ) comprising greater than 1% of the total CD62Lhi memory pool was counted . In each of the organs examined , earlyM contained 2–3 fold greater numbers of subpopulations than lateM cells ( Fig 3B ) . Similar patterns were seen for endogenous GP33 and GP276 tetramer positive memory cells ( S1C and S1D Fig ) , and for CD62Lhi earlyM and lateM P14 cells co-stained with CD27 , CD127 , CD43 , CxCr3 , and CCR5 ( S3B Fig ) . Taken together , these data indicate that the phenotype of CD62Lhi memory CD8 T cells continues to change while heterogeneity of memory CD8 T cells decreases with time after infection . Differences in phenotype among CD62Lhi earlyM and lateM P14 cells suggested that gene expression and functional changes could also occur with time . In order to determine the extent of changes in gene expression that occur with time within CD62Lhi memory subsets and to determine if genes regulating memory CD8 T cell functions were differently expressed , we examined the transcriptomes of CD62Lhi earlyM and lateM P14 cells . We detected 3 , 494 genes that were differentially expressed ( p<0 . 05 ) between CD62Lhi earlyM and lateM P14 cells , and Fig 4A shows a heat map of genes with significantly different expression between CD62Lhi earlyM and lateM P14 cells at fold differences >1 . 25 . To determine if gene families with function in T cell biology are differentially expressed between CD62Lhi earlyM and lateM P14 cells , we used DAVID bioinformatics resources [38] to assign biological functions and to group genes into function-related classes for genes with significantly different expression ( p<0 . 05 ) and fold >1 . 5 . This analysis revealed that mRNA expression for genes involved in many cellular processes are differentially expressed over time after infection in CD62Lhi memory cells ( Table 1 ) . Prominently among these changes , CD62Lhi lateM cells showed increased expression of cytokine receptors including Il2rb and IL15ra ( encoding IL-2Rβ and IL-15Rα respectively ) , and decreased expression of killer-cell lectin-like receptors including Klrg1 . Furthermore , lateM cells showed alterations in genes regulating cell cycle progression and ribosome biogenesis , suggesting that the proliferative potential of memory CD8 T cells may increase with time after infection not only at the population level ( Fig 2 ) , but also within defined subsets . Additionally , we used KEGG pathway analysis [39] to examine regulation of biological pathways in CD62Lhi earlyM and lateM cells . This analysis revealed that multiple pathways are differentially regulated with time in CD62Lhi memory cells including cell cycle and ribosome pathways ( Fig 4B ) . It is recognized that metabolic pathways are dynamically regulated during CD8 T cell responses , and both naïve and memory CD8 T cells catabolize fatty acid and utilize oxidative phosphorylation to meet their energy requirements [40–47] . The DAVID and KEGG analyses revealed that metabolism-related genes/pathways including metabolism of fatty acid , fructose and mannose metabolism , and oxidative phosphorylation were altered with time after infection in CD62Lhi memory CD8 T cells ( Table 1 and Fig 4B ) . Taken together , these data further indicate that transcriptomic alterations occur within CD62Lhi memory CD8 T cells over time after infection . Memory CD8 T cells rapidly respond to infection with the production of cytokines and the release of cytolytic molecules including perforin and granzymes [3] . While microarray data did not indicate differences in mRNA expression of effector molecules between resting earlyM and lateM P14 cells , this did not rule out the possibility that effector functions of CD62Lhi memory CD8 T cells change with time in response to Ag-stimulation . To examine if cytokine production by CD62Lhi memory CD8 T cells changes with time , Thy disparate earlyM and lateM P14 cells were mixed together and incubated with GP33-41 peptide for 5 h followed by ICS for detection of IFN-γ , TNF-α , IL-2 , and CD107a as a measure of degranulation . No differences in the production of any cytokines or degranulation were observed between earlyM and lateM CD62Lhi P14 cells ( Fig 5A and 5B ) . To determine if sensitivity to Ag changes with time in CD62Lhi memory CD8 T cells , we performed ICS as described above using decreasing concentrations of GP33-41 peptide . No differences in Ag sensitivity were detected between CD62Lhi earlyM and lateM P14 cells based upon functional avidity curves and the effective concentration of peptide required to induce 50% of cells to produce IFN-γ ( EC50 ) ( Fig 5C and 5D ) . Additionally , when earlyM and lateM P14 cells were incubated with GP33-41 peptide for decreasing lengths of time , no differences in the time required to produce IFN-γ , TNF-α , or IL-2 were observed ( Fig 5E ) , suggesting that there is a similarly poised state of earlyM and lateM CD62Lhi to produce cytokines upon Ag recognition . As suggested from the microarray , these data indicate that effector functions of CD62Lhi memory CD8 T cells including cytokine production , secretion of cytolytic molecules , Ag sensitivity , and time required to produce cytokines does not change with time after infection . To gain a broader appreciation for the poised state of earlyM compared to lateM CD62Lhi cells , we incubated CD62Lhi earlyM and lateM cells for 5 h in the presence or absence of GP33-41 peptide and examined expression of a number of effector molecules , surface markers , transcription factors , and cell cycle associated genes known to be dynamically regulated following Ag-encounter [41 , 48–50] . As we had previously noted by flow cytometry for expression of IFN-γ , TNF-α , and IL-2 ( Fig 5A and 5B , S4A Fig ) , mRNA expression of effector molecules was similar between CD62Lhi earlyM and lateM cells following incubation for 5 h with cognate Ag ( S5A Fig ) . While expression of some genes including Klrg1 and Cx3Cr1 were differently expressed between resting CD62Lhi earlyM and lateM cells ( S5B Fig ) as was indicated from the microarray data ( Table 1 ) , mRNA ( S5B Fig ) and surface protein ( S4B Fig ) expression of activation markers was similar for CD62Lhi earlyM and lateM cells following the 5 h incubation period , indicating that CD62Lhi earlyM and lateM cells are similarly activated following pathogen re-encounter . Additionally , mRNA ( S5C Fig ) and protein levels ( S4C Fig ) of transcription factors were regulated to a similar extent in CD62Lhi earlyM and lateM cells following 5 h peptide incubation . Interestingly , while the microarray data indicated that CD62Lhi lateM cells regulate expression of cell cycle related genes differently compared to earlyM cells ( Fig 4B and Table 1 ) , mRNA levels of cyclins and cyclin dependent kinases was simarly regulated in CD62Lhi earlyM and lateM cells following 5 h incubation with cognate Ag ( S5D Fig ) . Taken together , these data indicate that CD62Lhi earlyM and lateM cells display a similarly poised state to respond following Ag re-encounter . While mRNA expression of cell cycle genes following 5 h incubation with Ag indicated that genes regulating cell cycling were similarly regulated in CD62Lhi earlyM and lateM cells following Ag re-encounter ( S5D Fig ) , KEGG pathway analysis indicated that genes regulating cell cycle pathways are differentially expressed in CD62Lhi memory CD8 T cells with time ( Fig 4B and S6A Fig ) . Additionally , gene set enrichment analysis ( GSEA ) [51] comparing gene expression patterns of CD62Lhi earlyM and lateM P14 cells with existing gene sets revealed that genes highly expressed by CD62Lhi lateM P14 cells were enriched in gene sets involved in cell cycle pathways ( S6B Fig ) . Furthermore , while proliferation of CD62Lhi earlyM and lateM cells incubated with cognate Ag in vitro was similar after 5 h , a greater percentage of CD62Lhi lateM compared to earlyM cells proliferated following incubation with cognate Ag for 24 h ( S4D Fig ) . Taken together , these data suggested that CD62Lhi lateM CD8 T cells might possess enhanced abilities to undergo homeostatic proliferation and/or Ag-driven proliferative expansion following re-infection . Basal and homeostatic proliferation is partially dependent upon IL-15 [33–35] . IL-15 bound to IL-15Rα is trans-presented to CD8 T cells and signals through the common receptor γ chain ( γc , CD132 ) and IL-2Rβ ( CD122 ) [52] . mRNA expression of CD122 as determined from microarray data ( Fig 6A ) and surface expression of CD122 as detected by flow cytometry ( Figs 3A and 6B ) , increased with time in CD62Lhi memory P14 cells , suggesting that sensitivity to IL-15 could increase in CD62Lhi memory CD8 T cells with time after infection . To test this , Thy disparate earlyM and lateM P14 cells were labeled with CFSE , mixed and incubated with increasing concentrations of IL-15 , and dilution of CFSE was determined after 3 d . CD62Lhi lateM P14 cells proliferated to a greater extent in response to exogenous IL-15 , and CD62Lhi lateM cells were more sensitive to lower levels of IL-15 compared to CD62Lhi earlyM P14 cells ( Fig 6C ) . Endogenous GP33 and GP276 CD62Lhi lateM cells also displayed increased expression of CD122 compared to earlyM cells ( S1B Fig ) suggesting they possesses increased sensitivity to IL-15 compared to earlyM CD62Lhi cells . Taken together , this data indicates that sensitivity to IL-15 increases over time in CD62Lhi memory CD8 T cells . Differences in sensitivity to IL-15 suggested that CD62Lhi earlyM and lateM CD8 T cells could differ in their ability to undergo homeostatic proliferation . To test this , Thy disparate CD62Lhi earlyM and lateM P14 cells were sorted , mixed in equal numbers ( Fig 6D left panel ) , injected into lymphopenic Rag-/- mice ( 3x104 each ) , and the percentage of CD62Lhi earlyM and lateM cells was determined in the spleens of recipients 15 d after transfer . A higher percentage of lateM cells was found with the ratio of lateM to earlyM cells increasing approximately two-fold ( Fig 6D right panel ) , indicating that ability to undergo homeostatic proliferation increases with time in CD62Lhi memory CD8 T cells . Differences in cell cycle regulation also could indicate that the ability to undergo Ag-driven proliferation changes with time among CD62Lhi memory CD8 T cells . To test this , Thy disparate CD62Lhi earlyM and lateM P14 cells were sorted , mixed in equal numbers ( Fig 7A ) , and 1x104 of each were transferred into naïve C57BL/6 recipients followed by infection with LCMV 24 h later . An increased percentage of progeny were generated from CD62Lhi lateM P14 cells during the effector phase ( Fig 7B–7D ) , indicating that with time CD62Lhi memory CD8 T cells have an increased ability to undergo secondary expansion . Adoptive transfer of endogenous CD62Lhi GP276-specific earlyM and lateM cells ( S7A Fig ) also showed increased numbers of 2° effector cells generated from lateM cells compared to earlyM cells in PBL and spleen ( S7B and S7C Fig ) , and nearly 4 times the number of 2° effector cells generated from CD62Lhi lateM cells compared to earlyM cells were recovered in spleens 7 days post LCMV infection ( S7C Fig ) . Additionally , when progeny generated from CD62Lhi earlyM and lateM P14 cells were examined at a memory time point in PBL ( Fig 7C and 7D ) or in peripheral tissues and secondary lymphoid organs ( Fig 7E ) , a greater percentage of secondary memory cells were generated from CD62Lhi lateM compared to earlyM , indicating that with time , memory generation potential increases for CD62Lhi memory CD8 T cells . As suggested by the microarray data , these results indicate that with time after infection , the ability to undergo Ag-driven proliferation and generate secondary memory cells increases within CD62Lhi memory CD8 T cells . The goal of vaccination is establishment of memory populations that will provide increased protection upon infection , and studies have indicated that the quantity , quality , and localization of memory CD8 T cells required for protection differ depending upon the nature of the pathogen [6 , 8–10 , 53 , 54] . The functional differences that we observed suggested that CD62Lhi earlyM and lateM cells might provide differing levels of protection following infection . To determine if per cell protective capacity of CD62Lhi memory CD8 T cells against an acute systemic infection changes with time , CD62Lhi earlyM and lateM P14 cells were sorted , and 7x104 cells were transferred into naïve C57BL/6 recipients followed by infection with virulent LM expressing GP33 . Both CD62Lhi earlyM and lateM P14 cells provided protection , as significantly decreased ( p<0 . 05 ) colony forming units ( CFUs ) of LM were detected in spleens of recipient mice three days after infection compared to mice not receiving adoptive transfer . However , per cell protective capacity of CD62Lhi earlyM and lateM P14 cells did not differ ( Fig 8A ) . The most dramatic alteration in functional ability between CD62Lhi earlyM and lateM CD8 T cells that we observed was the ability to undergo Ag-driven proliferation , and some studies have indicated that the proliferative abilities of memory CD8 T cells are less important than localization and killing ability for providing protection from LM [8–10] . However , studies have shown that proliferative abilities of memory CD8 T cells are crucial for clearance of infection with LCMV clone-13 , which causes a chronic infection in mice [6 , 10 , 54] . To determine if per cell protective capacity of CD62Lhi memory CD8 T cells against a chronic infection changes with time , we sorted CD62Lhi earlyM and lateM P14 cells and transferred 5x104 cells into naïve C57BL/6 recipients followed by infection with LCMV clone-13 . Mice that received adoptive transfer of CD62Lhi lateM cells had reduced viral titers eight days following infection compared to mice not receiving transferred cells , and this level of protection was significantly greater than that provided by CD62Lhi earlyM cells ( Fig 8B ) . Increased protection provided by lateM CD62Lhi cells correlated with enhanced magnitudes of proliferative expansion as a greater percentage of progeny generated from CD62Lhi lateM compared to earlyM cells was found in the PBL ( Fig 8C ) , and higher numbers of progeny generated from CD62Lhi lateM were found in the spleens of recipient mice 8 days p . i . ( Fig 8D ) . Taken together , these data indicate that protection provided by CD62Lhi memory CD8 T cells changes with time after infection in a pathogen-dependent manner . Protection provided against infection is mediated by all memory cells present at the time of re-infection . When analyzed on the population level ( i . e . no sorting based on CD62L expression ) , lateM CD8 T cells proliferated to a greater extent following acute infection compared to earlyM populations ( Fig 2B ) , suggesting that protection provided by populations of memory CD8 T cells may also differ with time . To determine if per cell protective capacity of memory CD8 T cell populations changes with time , cells were isolated from the spleens of mice containing earlyM or lateM P14 cells , and 7x104 cells were transferred into naïve C57BL/6 recipients followed by infection with virulent LM expressing GP33 . Both earlyM and lateM populations provided protection against LM , but no difference in the level of protection was observed ( Fig 8E ) . To determine if protection provided by memory CD8 T cell populations against a chronic infection changes with time after infection , 5x104 earlyM or lateM cells were transferred into naïve C57BL/6 recipients followed by infection with LCMV clone-13 . lateM populations provided enhanced protection against chronic LCMV clone-13 infection compared to earlyM populations ( Fig 8F ) . Enhanced protection provided by lateM CD8 T cells also correlated with enhanced secondary expansion in PBL and greater numbers of P14 cells recovered from spleens of infected mice 8 days following infection ( Fig 8G and 8H ) . Taken together , these data indicate that the ability of memory CD8 T cells to provide protection against chronic infection increases with time after infection in a manner not due solely to shifts in Tem to Tcm subsets . Alterations in metabolic function including enhanced fatty acid oxidation , increased mitochondrial mass , and increased ability to perform oxidative phosphorylation have been shown to enhance memory CD8 T cell development and favor rapid recall responses following Ag re-encounter [44–47 , 55] . Additionally , memory CD8 T cells that displayed an enhanced ability to undergo oxidative phosphorylation and that possessed increased spare respiratory capacity ( SRC ) , which is the reserve ATP generation capacity of cellular mitochondria , proliferated to a greater extent and provided enhanced protection against chronic infection with LCMV clone-13 [54] . Functional annotation and KEGG pathway analysis of our microarray data revealed that genes regulating mitochondrial function and metabolic pathways including oxidative phosphorylation; fructose , mannose , and glucose metabolism; and fatty acid oxidation were differently regulated between CD62Lhi earlyM and lateM CD8 T cells ( Table 1 , Figs 4B and 9A ) . Additionally , GSEA analysis revealed that lateM CD62Lhi CD8 T cells were enriched in genes sets involved in oxidative phosphorylation ( Fig 9B ) . This suggested that CD62Lhi memory CD8 T cells might alter their metabolic programs with time after infection , which would impact their ability to proliferate and to provide protection against LCMV clone-13 infection . To determine if the metabolic function of CD62Lhi memory CD8 T cells is altered with time after infection , we sorted earlyM and lateM CD62Lhi memory CD8 T cells and performed extracellular flux analysis using a Seahorse bioanalyzer . Comparison of the basal oxygen consumption rate ( OCR ) , a measure of oxidative phosphorylation , to the extracellular acidification rate ( ECAR ) , a measure of aerobic glycolysis , revealed that compared to CD62Lhi earlyM cells , CD62Lhi lateM cells rely to a greater extent on oxidative phosphorylation for the generation of ATP ( Fig 9C ) . Due to the demands of cell sorting , we were unable to determine SRC of CD62Lhi earlyM and lateM cells , as they did not respond to the fluorocarbonyl cyanide phenylhydrazone ( FCCP ) inhibitor . However , lateM CD8 T cell populations also proliferated to a greater extent than earlyM populations and provided enhanced protection against LCMV clone-13 , and these populations could be isolated without using cell sorting . To examine if the metabolic function of CD8 T cell memory populations is altered with time after infection we performed extracellular flux analysis of isolated earlyM and lateM CD8 T cell populations . lateM cells displayed increased basal OCR levels compared to earlyM cells , and both cell types responded to metabolic inhibitors ( Fig 9D ) . Comparing the ratio of OCR to ECAR for earlyM and lateM populations reveled that lateM cells rely more heavily on oxidative phosphorylation for ATP production compared to earlyM populations ( Fig 9E ) . Analysis of the SRC , calculated as the highest OCR after addition of FCCP over the basal OCR , for earlyM and lateM revealed a trend for higher SRC in populations of lateM cells compared to earlyM cells ( Fig 9F ) . Taken together , these data indicate that mitochondrial function of memory CD8 T cells improves with time after infection in a manner that is not solely due to shifts in Tcm and Tem subsets . The enhanced mitochondrial function of lateM populations and CD62Lhi cells likely provides lateM cells with a metabolic advantage enabling robust proliferation and enhanced protection against chronic infection with LCMV clone-13 .
Protection provided by memory CD8 T cells is dependent upon their numbers , functional ability ( quality ) , and location at the time of infection [56] . We have shown that the quality of the circulating memory CD8 T cell population differs with time after infection in a manner not solely due to shifts in memory subset composition . Some functions of memory CD8 T cells analyzed on the population level , such as ability to produce IL-2 , increased with time after infection , but were no different in CD62Lhi memory cells early or late after infection . However , other qualitative aspects of memory CD8 T cells including proliferation in response to the homeostatic cytokine IL-15 or to Ag , and mitochondrial function increased with time after infection when both the memory CD8 T cell population , and defined CD62Lhi subsets were analyzed . Thus , while some alterations in the functional abilities of memory CD8 T cells with time after infection can be attributed to shifts in subset composition , other qualitative changes cannot be wholly attributed to shifts in subset composition . Interestingly , as a consequence of these functional changes , the protection provided by memory CD8 T cell populations and CD62Lhi memory CD8 T cells against a chronic viral infection increased over time . Importantly , our data suggests that the outcome vaccination schemes designed to elicit protective memory CD8 T cells will depend on the timing between booster immunizations , and on the timing of re-infection following vaccination . While circulating memory CD8 T cells are best suited to provide protection against systemic infections , tissue resident memory CD8 T cells provide a first line of defense against pathogens encountered in peripheral tissues [7 , 11–15] . While the longevity of Trm cells relative to circulating memory CD8 T cells is unclear at present , some studies have indicated that Trm cells remain in mice for at least 300 days following infection with vaccinia virus [12] , herpes virus [15] , or vesicular stomatitis virus [57] . However , other studies examining Trm formation following infection with influenza virus have indicated that the Trm CD8 T cell population wanes following infection [17] . Thus , longevity of Trm CD8 T cells may vary depending on factors including the nature of the primary infection and/or vaccination and the tissue of residence . Additionally , studies examining the longevity and phenotypic and functional changes that might occur in Trm CD8 T cells over time following infection will be complicated due to a lack of phenotypic markers that definitively identify Trm cells , as CD103 and CD69 , markers used to identify Trm cells , are not expressed on all Trm cells [7 , 58 , 59] . However , Trm CD8 T cells likely would play an important role in providing protection against human pathogens that infect at peripheral tissues , including HIV , herpes viruses , and influenza viruses . Therefore , determining the longevity of Trm cells and whether the phenotype , function , and protective abilities of Trm cells differs with time after infection , as we have shown for circulating memory CD8 T cells , is an important goal . While we have provided evidence that changes in memory CD8 T cell phenotype and function seen on the population level are not due solely to conversion to CD62Lhi cells with time after infection , a question still remains as to how the progressive changes in phenotype and function seen with time after infection in both memory CD8 T cell populations and within CD62Lhi memory CD8 T cells occurs . Our microarray data suggests that transcription of genes important for memory CD8 T cell function are differentially regulated in CD62Lhi memory CD8 T cells with time after infection , but it is unclear if differences in the transcriptional program with time after infection are due to 1 ) cell-intrinsic changes in gene regulation , or 2 ) a subset of memory CD8 T cells within the earlyM population selectively survives and comes to constitute the lateM pool . To ideally address these possibilities , earlyM CD8 T cell subsets would be transferred into naïve mice , and the phenotype , function , protective abilities , and transcriptional regulation of the transferred cells would be analyzed for the lateM cells derived from the transferred populations . However , due to the low number of CD62Lhi memory CD8 T cells present early following infection , loss of cells upon adoptive transfer , and further loss of cells upon isolation from recipient mice , these experiments are difficult to execute . We hope that improvements in cell isolation technology will allow us to perform these experiments in the future . We considered the possibility that transcription factors that facilitate memory CD8 T cell formation such as Tbet , Eomes , or Tcf1 [60–62] could differentially regulate survival of subsets of memory cells displaying phenotypic markers that were present in earlyM cells but not present in lateM cells as seen in S4 Fig . However , while expression of Eomes and Tcf1 differed between earlyM and lateM populations ( S8 Fig ) , we were unable to find conclusive evidence that expression of these transcription factors regulated survival of subsets within CD62Lhi earlyM and lateM cells . The rate at which phenotypic and functional changes of memory CD8 T cells occur following infection and/or vaccination is likely to be influenced by a number of factors . In this study we examined Tg memory P14 cells primarily localized within the spleen following acute infection with LCMV . Previous studies have indicated that the rate of acquisition of effector functions and expression of surface markers associated with central memory CD8 T cells is influenced by the number of transferred Tg T cells [24 , 63 , 64] . However , expression of CD62L following adoptive transfer of low numbers of Tg P14 cells as used in our study has been shown to be similar between Tg and endogenous Db-gp33 restricted CD8 T cells [24 , 63] . On the other hand , the rate of surface CD62L expression and ability to produce IL-2 has been shown to differ between endogenous T cell populations of different LCMV epitope specificities and among CD8 T cells localized within secondary lymphoid organs or peripheral tissues [24] . Additionally , the nature of the infecting pathogen and/or inflammation elicited during either infection or vaccination has been shown to influence the phenotype and function of memory CD8 T cells generated during the response . Infection with LCMV , vaccinia virus , or influenza virus leads to the formation of memory CD8 T cells with distinct phenotypic and functional qualities [65 , 66] , while administration of inflammation inducing toll-like receptor agonists during dendritic cell ( DC ) immunizations abrogates the rapid acquisition of memory characteristics seen during low-inflammatory DC immunizations [67–69] . Because of these considerations , the extent of changes that occur within the CD8 T cell memory population with time , and thus their functional and protective abilities during re-infection will likely depend upon conditions elicited during the primary infection and/or immunization . Therefore , vaccine design should include considerations of how the vaccine strategy may influence changes in memory CD8 T cells with time . Unlike mice housed in specific pathogen free facilities , humans are infected with many non-related pathogens , and co-infections or chronic infections could influence the development and/or differentiation of primary memory CD8 T cells , or the properties of already established memory CD8 T cells . A recent report showed that established chronic infections in mice influence the development and differentiation of primary memory CD8 T cells , but that the impact of chronic infections on pre-established primary memory CD8 T cells was less severe [70] . However , pre-established memory CD8 T cells examined in their study were generated 1 year prior to the chronic infection . It will be interesting to examine if differentiation of more recently established memory CD8 T cells is also minimally impacted by chronic or repeated unrelated infections . Studies have indicated that the quantity , quality , and localization of memory CD8 T cells required for protection differ depending upon the nature of the pathogen [6 , 8–10 , 53 , 54] . We found that the protective abilities of memory CD8 T cells changes with time in a pathogen-dependent manner . CD62Lhi central memory cells have been described as being more effective at providing protection against chronic infections due to localization within the lymph nodes and increased mitochondrial function leading to an enhanced ability to proliferate [6 , 10 , 54] . We showed that the ability of memory CD8 T cell populations and CD62Lhi central memory CD8 T cells to provide protection from LCMV clone-13 infection increases with time , and that this increased protection correlated with enhanced mitochondrial function and proliferative abilities following infection . Some studies [8–10] , however , have indicated that memory CD8 T cells with effector memory characteristics provide increased protection against acute infection with L . monocytogenes or localized infection with vaccinia virus , and the localization of the memory population to sites of infection is important in these instances . Therefore , increased protection provided by lateM cells likely does not apply to all infections . The numbers of memory CD8 T cells required to achieve protection against certain pathogens including Plasmodium species which cause malaria is quite high , and prime boost protocols have been established in order to achieve high numbers of memory CD8 T cells [29–31] . Our data indicate that higher numbers of memory CD8 T cells may be achieved through prime boost protocols by increasing the length of time between boosts . However , our study analyzed primary memory cells , and recent studies have indicated that the properties of memory CD8 T cells including magnitude of proliferative expansion , duration and degree of contraction , cytotoxicity , IL-2 production , basal proliferation and long-term survival , memory generation potential , lymph node homing , and transcriptome diversification change sequentially with each additional Ag encounter [71–73] . While little is known about how the number of Ag encounters influences the changes in memory CD8 T cell functions that occur with time after infection , studies indicate that the phenotype of memory CD8 T cells that have encountered Ag multiple times changes with time after infection , but at a slower rate than in primary memory CD8 T cells [25 , 28] . As with primary memory , changes with time in the properties of memory CD8 T cells that have encountered Ag more than once could influence their ability to provide protection against infection and/or affect the outcome of prime boost immunizations requiring multiple boosts . Our results firmly establish that memory CD8 T cells continue to change with time after infection . The results indicate that the function of memory CD8 T cells continues to change with time after infection , and that protection provided by memory CD8 T cells changes with time in a pathogen-dependent manner . Because of this , experimental investigation of memory CD8 T cell quality and/or protection following either infection or vaccination should include analysis of memory CD8 T cells at multiple time points following the infection and/or vaccination .
All experimental procedures utilizing mice were approved by the University of Iowa Animal Care and Use Committee under the ACURF protocol number 1202050 . The experiments performed in this study were done under strict accordance to the Office of Laboratory Animal Welfare guidelines and the PHS Policy on Humane Care and Use of Laboratory Animals . C57BL/6 Thy1 . 2 mice were obtained from the National Cancer Institute ( Frederick , MD ) . B6/SJL ( CD45 . 1 ) , Rag-/- mice , and P14 mice were bred at the University of Iowa ( Iowa City , IA ) . The Armstrong strain of LCMV , the clone-13 strain of LCMV , Vaccinia Virus expressing the GP33 epitope ( VacV ) , attenuated actA- deficient Listeria monocytogenes expressing the GP33 epitope [31] , and virulent Listeria monocytogenes strain 1043S expressing the GP33 epitope were grown and quantified as previously described [10 , 27] . For generation of earlyM and lateM P14 cells , P14 CD8 T cells ( specific for the LCMV GP33-41 epitope ) were isolated from the peripheral blood of young Thy1 . 1/1 . 1 or Thy1 . 1/1 . 2 P14 mice . Contaminating memory phenotype ( CD11ahi/CD44hi ) P14 cells were always <5% . 5x103 P14 cells were transferred retro-orbitally into 6–12 week old naïve C57BL/6 mice , and recipients were infected 24 h later intraperitoneally ( i . p . ) with 2x105 plaque forming units ( PFU ) of LCMV Armstrong ( LCMV ) . All earlyM analysis was done between 30–45 days after infection and lateM analysis was done 8+ months after infection . For co-transfer of earlyM and lateM P14 cells , P14 cells were isolated from the spleens of mice containing Thy disparate earlyM and lateM P14 cells , mixed at a 1:1 ratio , and 2x104 of each were transferred retro-orbitally into naïve C57BL/6 mice followed 24 hours later by i . p . injection of 2x105 PFU of LCMV or 3x106 PFU of VacV , or by intravenous ( i . v . ) injection of 5x106 colony forming units ( CFUs ) of LM . For co-transfer of CD62Lhi earlyM and lateM P14 cells , P14 cells were isolated from the spleens of mice containing Thy disparate earlyM and lateM P14 cells , and cells were surface stained for Thy1 . 1 , CD8 , and CD62L ( eBioscience ) and sorted using a BD FACSAria II ( BD Biosciences ) . Sorted cells were mixed at a 1:1 ratio and 1x104 of each were transferred retro-orbitally into naïve C57BL/6 mice followed 24 hours later by i . p . injection of 2x105 PFU of LCMV . Sorted cells were >95% pure . Input ratios of earlyM and lateM P14 cells were confirmed by flow cytometry before adoptive transfer . For adoptive transfer of CD62Lhi endogenous GP276 cells , splenocytes of earlyM and lateM C57BL/6 ( CD45 . 2 ) mice were stained with PE-anti-CD8 antibodies and purified with anti-PE magnetic bead sorting using standard AutoMacs protocols . Cells were then stained with CD62L , and CD62Lhi cells were sorted using a BD FACSAria II ( BD Biosciences ) . Following sorting , cells were stained with GP276 tetramer to determine the percentage of endogenous GP276 memory cells in the sorted CD62Lhi CD8 T cell population . 2 . 5x103 endogenous CD62Lhi earlyM or lateM GP276 tetramer positive cells were then transferred into CD45 . 1 C57/SJL mice followed 24 hours later by i . p . injection of 2x105 PFU of LCMV . earlyM and lateM CD8 T cell responses were quantified in peripheral blood by collecting blood via retro-orbital puncture . Red blood cells were lysed with ACK , and P14 cells were surface stained for Thy1 . 1 , Thy1 . 2 , and CD8 ( eBioscience ) , and endogenous memory cells were surface stained for Thy1 . 1 , CD45 . 2 , CD8 ( eBioscience ) , and GP276 tetramer . Cells were acquired on a FACSCalibur instrument ( BD Biosciences ) , and Thy expression was used to distinguish between earlyM and lateM P14 cells . For isolation of lymphocytes from tissues , anesthetized mice were perfused through the left ventricle with PBS and single-cell suspensions from the lung , spleen , liver , and inguinal lymph nodes were prepared as previously described [27] . Surface marker expression and heterogeneity among earlyM and lateM P14 cells was determined by 8 color staining of isolated lymphocytes for Thy1 . 1 , CD8 , CD27 , CD122 , KLRG1 , CD11b ( eBioscience ) , CD62L , and CD127 ( biolegend ) , or 8 color staining of isolated lymphocytes for Thy1 . 1 , CD8 , CD27 ( eBioscience ) , CxCr3 , CD127 , CCR5 , and CD43 ( biolegend ) . Surface marker expression and heterogeneity among endogenous GP33 and GP276 earlyM and lateM cells was determined by 8 color staining of isolated splenocytes for GP33 or GP276 tetramer , Thy1 . 1 , CD8 , CD27 , CD122 , KLRG1 ( eBioscience ) , CD62L and CD127 ( biolegend ) . Cells were acquired on a LSR II instrument ( BD Biosciences ) , and gates were set using fluorescence minus one ( FMO ) staining . Ex vivo cytokine detection was determined as previously described [74] by mixing splenocytes containing earlyM and lateM P14 cells and incubating with 200nM GP33-41 peptide at 37°C for 5 hours in the presence of Brefeldin A ( BFA ) for 5 hours or 1 hour . To prevent CD62L cleavage , cells were pre-incubated for ½ hour in the presence of 100μM TAPI-2 ( Peptides International ) as previously described [75] . To determine functional avidity , splenocytes were incubated as described above in the presence of the indicated concentrations of GP33-41 peptide as described previously [76] . To determine the time required to produce IFN-γ , TNF-α , or IL-2 , splenocytes containing earlyM and lateM P14 cells were mixed together and incubated as described above in the presence of 200nM GP33-41 peptide for the indicated periods of time . Cells were surface stained for Thy1 . 1 , Thy1 . 2 , CD8 , and CD62L ( eBioscience ) , then permeabalized and stained intracellularly for IFN-γ , TNF-α , or IL-2 production . Some samples were also surface stained for expression of CD25 , CD69 , and CD122 . For detection of degranulation , cells were pre-incubated with TAPI-2 for ½ hour then in the presence of 200nM GP33-41 peptide plus monensin and anti-CD107a antibodies ( BD Biosciences ) for 5 hours at 37°C prior to surface staining as previously described [76] . For detection of cycling , cells were pre-incubated with TAPI-2 for ½ hour then in the presence of 200nM GP33-41 for 5–24 hours at 37°C . Cells were then surface stained for CD8 , Thy1 . 1 , Thy1 . 2 , and CD62L ( eBioscience ) , then permeabilized using a Foxp3 staining kit ( ebioscience ) and stained intercellularly with Abs against KI67 ( BD Pharmingen ) . Cells were acquired on a FACSCanto instrument ( BD Biosciences ) . For detection of polyfunctional cytokine production , earlyM and lateM P14 cells from spleens were incubated in the presence of 200nM GP33-41 peptide as described above . Cells were surface stained for CD8 , Thy1 . 1 , and CD62L ( eBioscience ) then permeabilized and stained intracellularly for IFN-γ , TNF-α , and IL-2 production . Cells were acquired on a LSR II instrument ( BD Biosciences ) , and gates were set using cells incubated in the absence of GP33-41 peptide . To determine expression of Eomes and TCF1 , splenocytes were surface stained for CD8 , Thy1 . 1 , and CD62L , permeabilized using a Foxp3 staining kit ( ebioscience ) , and stained intercellularly with Abs against Eomes ( ebioscience ) or TCF1 ( Cell signaling ) . For detection of basal proliferation , mice were i . p . injected with 2mg BrdU and given 0 . 8 mg/mL BrdU in drinking water for an additional 8 days . P14 cells isolated from peripheral blood were surface stained for CD8 and Thy1 . 1 ( eBioscience ) followed by fixation and permeabilization procedures as recommended in the BrdU flow kit ( BD Biosciences ) . Anti-BrdU mAb ( eBioscience ) was used for intracellular staining to detect BrdU uptake . Cells were acquired on a FACSCalibur instrument ( BD Biosciences ) . For determination of responsiveness to IL-15 , splenocytes containing earlyM and lateM P14 cells were mixed together , washed three times in PBS , and CFSE labeled by incubating 107 splenocytes/mL in room temperature PBS for 15 minutes at 37°C in the presence of 5μM CFSE . CFSE-labeled cells were incubated on ice for 5 minutes with 1mL of fetal calf serum ( FCS ) and washed three times with RPMI 1640 containing 10% fetal calf serum . CFSE labeled cells were incubated for 3 days at 37°C in the presence or absence of the indicated concentrations of recombinant mouse IL-15 ( biolegend ) . Cells were surface stained for Thy1 . 1 , Thy1 . 2 , CD8 , and CD62L ( eBioscience ) and acquired on a FACSCanto instrument ( BD Biosciences ) . Cells incubated without IL-15 were used to set gates for CFSE dilution . For detection of homeostatic proliferation in Rag-/- mice , P14 cells were isolated from spleens of mice containing Thy disparate earlyM and lateM P14 cells , and cells were surface stained for Thy1 . 1 , CD8 , and CD62L ( eBioscience ) and sorted using a BD FACSAria II ( BD Biosciences ) . Sorted cells were mixed at a 1:1 ratio and 3x104 of each were transferred retro-orbitally into Rag-/- mice . Sorted cells were >95% pure . Input ratios of earlyM and lateM P14 cells were confirmed by flow cytometry before adoptive transfer . Rag-/- mice were sacrificed on d15 after transfer , and ratios of earlyM and lateM P14 cells in spleens were determined by surface staining for Thy1 . 1 , Thy1 . 2 , and CD8 . Cells were run on a FACSCalibur instrument ( BD Biosciences ) , and earlyM and lateM P14 cells were distinguished based on Thy disparity . Total splenocytes from mice containing earlyM or lateM P14 cells were stained with PE-anti-Thy1 . 1 antibodies and purified with anti-PE magnetic bead sorting with standard AutoMacs protocols . CD62Lhi cells were further surface stained for CD8 and CD62L ( ebioscience ) and sorted using a BD FACSAria II ( BD Biosciences ) . For determination of protection based on bacterial clearance , 7x104 PE-selected earlyM or lateM populations or sorted CD62Lhi earlyM or lateM P14 cells were transferred retro-orbitally into naïve C57BL/6 mice followed 24 hours later by i . v . injection of 1x105 CFU of virulent Listeria monocytogenes expressing the GP33 epitope . Three days after infection , spleens were harvested and placed in sterile deionized water containing 0 . 2% IGEPAL and disrupted using a tissue homogenizer . Samples were plated on tryptic soy broth ( TSB ) -agar plates containing streptomycin and incubated at 37°C for 24 hours , then CFUs were counted . For determination of protection based on viral clearance , 5x104 PE selected earlyM or lateM populations or sorted CD62Lhi earlyM or lateM P14 cells were transferred retro-orbitally into naïve C57BL/6 mice followed 24 hours later by i . v . infection of 2x106 PFU of LCMV clone-13 . 8 days after infection spleens were obtained and homogenized , and viral titers were quantified with standard plaque assaying on VERO cells as previously described [10] . Sorted cells were >95% pure . Control naïve mice did not receive adoptive transfer of P14 cells . CD62Lhi earlyM and lateM P14 cells were isolated from spleens and cells were surface stained for Thy1 . 1 , CD8 , and CD62L ( eBioscience ) and sorted using a BD FACSAria II ( BD Biosciences ) . Samples from three individual mice were obtained for each group , and sorted cells were >95% pure . RNA was extracted using the RNEasy Kit ( QIAGEN ) , and 1-5ng of mRNA was used for microarray analysis . RNA quality was assessed using the Agilent Model 2100 Bioanalyzer . mRNA for the microarray was processed using the NuGEN WT-Ovation Pico RNA Amplification System along with the NuGEN WT-Ovation Exon Module . Samples were hybridized and loaded onto Affymetrix GeneChip Mouse 1 . 0 ST arrays . Arrays were scanned with the Affymetrix Model 7G upgraded scanner , and data were collected using the GeneChip Operating Software . Data from the Affymetrix Mouse Exon 1 . 0 ST arrays were first quantile normalized and median polished using Robust Multichip Average background correction with log2 adjusted values . Probe sets for exons were then summarized for a specific gene using the median value . After obtaining log2 expression values for genes , significance testing was performed using analysis of variance ( ANOVA ) . Functional assignment of genes was performed using the “Functional Annotation Tools” in DAVID bioinformatics resources ( https://david . ncifcrf . gov ) following recommended protocols [38] . Enrichment of genes in known pathways was analyzed using the KEGG pathway tool , and GSEA was performed as described [51 , 77] . The microarray data were deposited in the NCBI Gene Expression Omnibus with the accession number GSE63307 . Total splenocytes from mice containing earlyM or lateM P14 cells were stained with PE-anti-Thy1 . 1 antibodies and purified with anti-PE magnetic bead sorting with standard AutoMacs protocols . CD62Lhi cells were further surface stained for CD8 and CD62L ( ebioscience ) and sorted using a BD FACSAria II ( BD Biosciences ) . Sorted CD62Lhi earlyM and lateM cells were then incubated with or without 200nM GP33-41 peptide at 37°C for 5 hours . Total RNA was reverse-transcribed using a QuantiTech Reverse Transcription Kit ( Qiagen ) . The resulting cDNA was analyzed for expression of different genes by quantitative PCR using SYBR Advantage qPCR pre-mix ( Clontech ) on an ABI 7300 Real Time PCR System ( Applied Biosystems ) . Relative gene expression levels in each sample were normalized to that of a housekeeping gene , hypoxanthine phosphoribosyltranserase 1 ( Hprt1 ) [62] . The primers used in quantitative RT-PCR were as follows: Ifng: 5’-GCGTCATTGAATCACACCTG and 3’-TGAGCTCATTGAATGCTTGG; Tnfa: 5’-TAGCCCACGTCGTAGCAAAC and 3’-GCAGCCTTGTCCCTTGAAGA; Il2: 5’-AACCTGAAACTCCCCAGGAT and 3’-CGCAGAGGTCCAAGTTCATC; Xcl1: 5’-ATGGGTTGTGGAAGGTGTGG and 3’-TGATCGCTGCTTTCACCCAT; Ccl3: 5’-CATATGGAGCTGACACCCCG and 3’-GTCAGGAAAATGACACCTGGC; Ccl5: 5’-GACAGCACATGCATCTCCCA and 3’-GTGTCCGAGCCATATGGTGA; Fasl: 5’-GCAGAAGGAACTGGCAGAAC and 3’-TTAAATGGGCCACACTCCTC; Klrg1: 5’-TCCTCTGGACGAGGAATGGT and 3’-ACAGCTTCACTCCCTGGTTG; Il2ra: 5’-GGTGCATAGACTGTGTTGGC and 3’-GCAAGAGAGGTTTCCGAAGAC; Cd69: 5’-ACATCTGGAGAGAGGGCAGA and 3’-AAGGACGTGATGAGGACCAC; Ccr5: 5’-CCCCTACAAGAGACTCTGGCTC and 3’-TTTTGGCAGGGTGCTGACAT; Il12rb2: 5’-GTGTCTGCAGCCAACTCAAA and 3’-AGGCTGCCAGGTCACTAGAA; Il18rap: 5’-GCAGGCTTACTCACCATTTCA and 3’-GCTTGTGCATCTTTATCCACGG; Cx3cr1: 5’-AAGTTCCCTTCCCATCTGCT and 3’-CGAGGACCACCAACAGATTT; Tbx21: 5’-TCAACCAGCACCAGACAGAG and 3’-CCACATCCACAAACATCCTG; Eomes: 5’-GGAAGTGACAGAGGACGGTG and 3’-AGCCGTGTACATGGAATCGT; Tcf7: 5’-CAATCTGCTCATGCCCTACC and 3’-CTTGCTTCTGGCTGATGTCC; Prdm1: 5’-CCTGCCAACCAGGAACTTCT and 3’-GTTGCTTTCCGTTTGTGTGAGA; Foxm1: 5’-CGAGCACTTGGAATCACAGC and 3’-GGATGGGCACCAGGTATGAG; Id2: 5’-CATCAGCATCCTGTCCTTGC and 3’-GTGTTCTCCTGGTGAAATGG; Id3: 5’-TGATCTCCAAGGACAAGAGGA and 3’-TGAAGAGGGCTGGGTTAAGA; Bcl2: 5’-GGAGGCTGGGATGCCTTTGT and 3’-TGCACCCAGAGTGATGCAG; Bcl6: 5’-CCTGAGGGAAGGCAATATCA and 3’-CGGCTGTTCAGGAACTCTTC; Foxo3: 5’-CTCATGGATGCTGACGGGTT and 3’-CGTCAGTTTGAGGGTCTGCT; Stat3: 5’-TGGTGTCCAGTTTACCACGA and 3’-TGTTCGTGCCCAGAATGTTA; Stat4: 5’-TTTTGACGCTGCAAGAAATG and 3’-TCCAGTCCTGCAGCTCTTCT; Myc: 5’-GTACCTCGTCCGATTCCACG and 3’-GCCTCTTCTCCACAGACACC; Ccnd2: 5’-TCAGTGTGGGTGATCTTGGC and 3’-CAGACCTTCATCGCTCTGTG; Ccnd3: 5’-GGACACTCGCTTTGTTTGGG and 3’-AGCATTTCAGGGCGAGCTTA; Ccne1: 5’-GTGGAGCTTATAGACTTCGCAC and 3’-ACTTACCTGAGAGATGAGCACT; Ccne2: 5’-AGAGTCGATGGCTAGAATGC and 3’-TGTCCAGTAACAGTCATCTCCT; Ccna2: 5’-GGTGAAGGCAGGCTGTTTAC and 3’-AGAAGCTCAAGACTCGACGG; Ccnb1: 5’-CCTGAGCCTGAACCTGAACT and 3’-ACGTCACTCACTGCAAGGAT; Ccnb2: 5’-GCAGAGCAGAGCATCAGAGA and 3’-CAGCCTCTGTGAAACCAGTG; Cdk1: 5’-TCAAGTCTCTGTGAAGAACTCG and 3’-TCCATGGACCTCAAGAAGTACC; Cdk2: 5’-CAATGCAGAGGGGTCCATCA and 3’-ACACACTAGGTGCATTTCAGC; Cdk4: 5’- CAGGTAGGAGTGCTGCAGG and 3’-AGTCAGTGGTGCCAGAGATG; Cdk5: 5’-GGATCTTCCGACTGCTAGGG and 3’-GCTGCACAGGGTTACACTTC; Cdk6: 5’-GCATCGTGATCTGAAACCGC and 3’-GTGACGACCACCGAGGTAAG . Both earlyM and lateM populations and CD62Lhi subsets were analyzed for metabolic function . Populations of earlyM and lateM cells were isolated from total splenocytes from mice containing earlyM or lateM P14 cells by staining with PE-anti-Thy1 . 1 antibodies and purifying with anti-PE magnetic bead sorting using standard AutoMacs protocols . Additional cells were sorted for CD62Lhi subsets after AutoMacs purification by surface staining for CD8 and CD62L ( ebioscience ) and sorting using a BD FACSAria II ( BD Biosciences ) . 2x105 purified earlyM or lateM populations or CD62Lhi cells were plated in XF media , and oxygen consumption rates ( OCR ) and extracellular acidification rates ( ECAR ) were measured under basal conditions and in response to 1 μM oligomycin , 1 . 5μM fluorocarbonyl cyanide phenylhydrazone ( FCCP ) , and 0 . 5 μM rotenone + 1μM antimycin with the XF-96 Extracellular Flux Analyzer ( Seahorse Bioscience ) [46] . Sell ( CD62L ) ID:20343 , CD27 ID:21940 , CCR7 ID:12775 , Thy1 ID:21838 , IL7R ( CD127 ) ID:16197 , IL2Rb ( CD122 ) ID:16185 , KLRG1 ID:50928 , IFNg ID:15978 , TNFa ID:21926 , Lamp1 ( CD107a ) ID:16783 , IL2 ID:16183 , IL15 ID:16168 , Itgam ( CD11b ) ID:16409 , CxCr3 ID:12766 , CCR5 ID:12774 , IL15ra ID:16169 , PRF1 ( perforin ) ID:18646 , GZMB ( granzymeB ) ID:3002 , Il2rg ID:16186 , Itgal ( CD11b ) ID:16408 , CD44 ID:12505 , Anapc5 ID:59008 , Bub1 ID:12235 , Ccnb2 ID:12442 , Ccne2 ID:12448 , Ccnh ID:66671 , Cdc7 ID:12545 , Cdk1 ID:12534 , Cdk4 ID:12567 , Cdkn2c ID:12580 , Mad2l1 ID:56150 , Orc6l ID:56452 , Rb1 ID:19645 , Skp1a ID:21402 , Ttk ID:22137 , Abl1 ID:11350 , Ccnd3-ps ID:626000 , Cdc25b ID:12531 , Gadd45g ID: 23882 , Zbtb17 ID:22642 , Atp5e ID:67126 , Atp5f1 ID:11950 , Atp5g2 ID:67942 , Atp5l ID:27425 , Atp5o ID:28080 , Atp6v0c ID:11984 , Atp6v1e1 ID:11973 , Atp6v1f ID:66144 , Atpgv1g1 ID:66290 , Cox4i1 ID:12857 , Cox6a1 ID:12861 , Cox7b ID:66142 , Ndufa2 ID:17991 , Ndufa11 ID:69875 , Ndufb6 ID:230075 , Ndufb8 ID:67264 , Ndufc2 ID:68197 , Ndufv1 ID:17995 , Ndufv2 ID:72900 , Sdhb ID:67680 , Sdhc ID:66052 , Uqcr1 ID:7384 , Uqcrq ID:22272 , Atp4b ID:11945 , Atp6v0e2 ID:76252 , IL2ra ID:16184 , CD69 ID:12515 , IL12rb2 ID:16162 , IL18rap ID:16174 , Cx3Cr1 ID:13051 , GZMA ( granzymeA ) ID:14938 , GZMK ( granzymeK ) ID:14945 , Xcl1 ID:16963 , Ccl3 ID:20302 , Ccl5 ID:20304 , Fasl ID:14103 , Tbx21 ( Tbet ) ID:57765 , Eomes ID:13813 , Tcf7 ( Tcf1 ) ID:21414 , Prdm1 ( blimp1 ) ID:12142 , Foxm1 ID:14235 , Id2 ID:15902 , Id3 ID:15903 , Bcl2 ID:12043 , Bcl6 ID:12053 , Foxo3 ID:56484 , Stat3 ID:20848 , Stat4 ID:20849 , Myc ID:17869 , Ccnd2 ID:12444 , Ccnd3 ID:12445 , Ccne1 ID:12447 , Ccna2 ID: 12428 , Ccnb1 ID: 268697 , Cdk2 ID:12566 , Cdk5 ID:12568 , Cdk6 ID:12571 | Following infection or vaccination , memory CD8 T cells persist at higher numbers and have enhanced functional abilities compared to naïve cells , providing immune hosts with increased protection from viral , bacterial , or parasitic infection . Protection provided by memory CD8 T cells depends on the numbers , quality ( functional abilities ) , and location of cells present at the time of re-infection . While memory CD8 T cells can be maintained for great lengths of time , how time influences qualitative properties of these cells remains largely unknown . We show that the phenotype and functions of circulating memory CD8 T cells , including cytokine production , proliferation , and mitochondrial function following re-infection improves with time after infection . We also show that changes in function are not due solely to changes in subset composition of the memory pool . Importantly , due to enhanced proliferative and metabolic abilities , memory CD8 T cells analyzed late after infection were more protective against a chronic viral infection . Our study shows that the properties of memory CD8 T cells continue to change with time , and that the protective outcome of vaccination may depend on the timing of re-infection relative to the initial immunization . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Phenotypic and Functional Alterations in Circulating Memory CD8 T Cells with Time after Primary Infection |
Here we report the first quantitative analysis of spiking activity in human early visual cortex . We recorded multi-unit activity from two electrodes in area V2/V3 of a human patient implanted with depth electrodes as part of her treatment for epilepsy . We observed well-localized multi-unit receptive fields with tunings for contrast , orientation , spatial frequency , and size , similar to those reported in the macaque . We also observed pronounced gamma oscillations in the local-field potential that could be used to estimate the underlying spiking response properties . Spiking responses were modulated by visual context and attention . We observed orientation-tuned surround suppression: responses were suppressed by image regions with a uniform orientation and enhanced by orientation contrast . Additionally , responses were enhanced on regions that perceptually segregated from the background , indicating that neurons in the human visual cortex are sensitive to figure-ground structure . Spiking responses were also modulated by object-based attention . When the patient mentally traced a curve through the neurons’ receptive fields , the accompanying shift of attention enhanced neuronal activity . These results demonstrate that the tuning properties of cells in the human early visual cortex are similar to those in the macaque and that responses can be modulated by both contextual factors and behavioral relevance . Our results , therefore , imply that the macaque visual system is an excellent model for the human visual cortex .
The early visual cortex consists of three areas , V1 , V2 , and V3 , which provide a retinotopic map of the visual field . Our knowledge of the properties of neurons in early visual cortex derives largely from electrophysiological studies of animal models , including the cat , macaque monkey , and more recently , the mouse . The pioneering work of Hubel and Wiesel revealed that cells in early visual areas respond to visual stimuli in their receptive field , a circumscribed region of the retina . Visual cortical neurons are typically tuned for orientation [1] and spatial frequency [2] and give saturating responses when the contrast of the stimulus increases [3] . Later studies revealed that the neuronal responses in early visual cortex can also be modified by the context set by image elements outside the neurons’ receptive field . For example , texture-defined figures elicit stronger responses than textured backgrounds if the receptive field stimulus is held constant [4] , and cognitive factors such as visual attention also influence the neuronal responses [5] . The usefulness of these data for our understanding of human vision depends on the similarities and differences between the animal models and the human [6] . So far , the comparison between animals and humans had to rely largely on post-mortem examinations to study the anatomy [7] and on indirect methods to measure brain activity such as functional magnetic resonance imaging ( fMRI ) [8] , electroencephalography ( EEG ) [9] , and magnetoencephalography ( MEG ) [10] , with subdural electrocorticography ( ECoG ) as the most direct , yet invasive method [11] . Quantitative descriptions of the activity profiles of cells in early human visual cortex have been lacking . Early studies have reported visually driven spiking activity from visual cortex neurons ( not localized to a particular area ) , but did not study them in great detail or quantify the receptive-field properties [12 , 13] . In this study , we report the properties of spiking activity recorded using microwires implanted in the occipital cortex of a patient during diagnostic surgery , part of her treatment for epilepsy . Most previous studies with microwires have targeted the medial temporal lobe of epileptic patients because this brain region is often implicated in the generation of epilepsy ( e . g . , [14–16] ) . Such recordings in visual cortex are much rarer as this region is almost never implicated in epileptogenesis . Here we report data from only two electrodes , and it is unlikely that we will be able to record more neurons from the same brain region in the near future . We measured the activity of neurons at these two electrodes in detail because the recordings were stable across a number of days . We could , therefore , for the first time to our knowledge , examine the tuning properties of the neurons in early visual cortex and explore how their activity is modulated by context and attention . We also recorded the local field potential ( LFP ) from the microwires , as recent data from patients implanted with ECoG grids suggest that the LFP can provide a first approximation of the tuning of spikes in visual cortex [11 , 17] . Our results demonstrate that spiking activity in human visual cortex shares many properties with that in macaque cortex , such as orientation and spatial frequency tuning , contrast saturation and surround suppression . We demonstrate , furthermore , that the spiking responses in early human visual cortex are enhanced by figure-ground segregation and object-based attention .
The early visual areas were mapped using standard retinotopic mapping techniques after explantation of the electrodes ( Materials and Methods ) . We observed a normal retinotopic representation in both hemispheres ( Fig 1B shows the left hemisphere ) . The positions of the microwires were judged from a post-implantation CT scan that was co-registered with a structural MRI image ( Fig 1B and 1C ) . The CT scan was used to visualize the macrocontacts of the electrode . The microcontacts were not visible on the CT scan but were situated between the macrocontacts ( Fig 1A ) . Electrodes E0–E3 were situated in or near the tip of the calcarine sulcus , most likely at very eccentric locations of V1 , and electrodes E4 and E5 were situated in white matter . We obtained multi-unit recordings from electrodes E6 and E7 , which were situated close to the representation of the horizontal meridian . Based on the location of the macroelectrodes , E6 and E7 were most likely located in V3 ( Fig 1C ) , but we shall refer to their location as V2/V3 throughout this report because previous studies in monkeys reported that it is difficult to unambiguously assign electrodes sites to V2 or V3 if they are situated close to the representation of the horizontal meridian [18] . In our electrophysiological experiments we used an eye tracker and aborted trials whenever the patient’s gaze fell outside a 2° radius window centered on the fixation point . We first localized the multi-unit receptive fields ( RFs ) of electrodes E6 and E7 using a 1° × 1° checkerboard ( check size: 0 . 33° ) presented at each location of an 11° × 11° grid . The RFs were both located at an eccentricity of 10°–11° , close to the horizontal meridian ( Fig 2A and 2E ) . We determined the average MUA responses at each location , in a 0 . 05–0 . 3 s window . We then fitted a 2D Gaussian and estimated the RF size as the full width at half maximum ( FWHM ) ( Materials and Methods ) . Neurons at E6 had a RF-size of 4 . 4° and neurons at E7 had a slightly smaller RF of 4 . 0° . We observed similar RF properties when examining the thresholded multi-unit signal ( S2A Fig ) . Previous studies have demonstrated that RF sizes of single-units in macaque V2/V3 at 10° eccentricity have an average value of 2°–3 . 5° ( V2: refs [19 , 20] ) or 3°–4 . 5° ( V3: refs [19 , 21] ) and multi-unit RFs in V2/V3 a size between 2°–3° ( V2: ref [22] ) and 4°–5° ( V3: refs [23–25] ) . The RF sizes in human V2/V3 are therefore in good agreement with the results from macaque V3 , increasing the likelihood that E6 and E7 were situated in V3 . A number of previous studies measured RFs in human visual cortex with the evoked potential or with the increase in power within the gamma band of intracranial ECoG recordings ( e . g . , [11 , 26] ) . To measure the influence of stimulus presentation on gamma power in the LFP , we computed the increase in power per frequency bin in a window from 100–250 ms after stimulus onset relative to the pre-stimulus epoch ( 150–0 ms before the stimulus ) . We observed that checks flashed in the RF caused a broad-band increase in gamma power ( 40–140 Hz ) ( Fig 2D and 2H ) . We used the increase in LFP gamma power to map the RFs . We observed clear RFs for E6 and E7 ( Fig 2C and 2G ) , which overlapped with the MUA-RFs but were , to our surprise , slightly smaller ( FWHM: E6 , 3 . 7°; E7 , 2 . 9° ) . Unlike previous reports [11] , we could not measure clear RFs using the average event-related potential of the LFP ( S3 Fig ) . The latency of the multi-unit response elicited by luminance flashes in the RF was 64 ms for E6 and 57 ms for E7 ( Fig 2B and 2F ) . These latencies are within the range of response latencies of single-units in macaque V2 and V3 [27] and also within the range of latencies of evoked potentials measured with ECoG grids [11] . Many neurons in the early visual cortices of mice , cats , and monkeys are tuned for the orientation , direction , spatial frequency , contrast , and size of visual stimuli . Studies using fMRI [8 , 28 , 29] have suggested that human visual cortex shares similar tuning characteristics , but spiking data is lacking . We therefore studied the tuning of neurons at electrodes E6 and E7 by presenting drifting sine-wave gratings ( see Materials and Methods ) in the RF while the patient maintained gaze at the fixation point . We assessed orientation tuning strength using 1 –circular variance ( 1-CircVar ) ( Materials and Methods ) . This measure ranges from 0 ( no tuning ) to 1 ( maximum tuning ) and is more robust to noise than other measures such as the orientation tuning index [30] . The spiking activity of the neurons at electrode E7 was significantly tuned for orientation ( Preferred orientation = 171 . 4° , i . e . , close to vertical , 1-CircVar = 0 . 10 , p < 0 . 001 , Bootstrap test ) , but not for motion direction ( 1-CircVarDir = 0 . 02 , p = 0 . 29 , Bootstrap test ) ( Fig 3A and 3B ) . The thresholded multi-unit signal from E7 ( MUAt ) was also tuned for orientation ( S2B Fig ) . The tuning width of the MUA signal was broad ( half width at half maximum [HWHM] = 58° ) but still within the range of single units in V2 and V3 of monkeys [31 , 32] . MUA from E6 was not significantly tuned to orientation ( p = 0 . 38 ) . We measured orientation tuning on two separate days , allowing us to assess the reliability of the signal across sessions ( S4 Fig ) . We found that signals from E7 were significantly tuned for orientation in both sessions ( Session 1: 1-CircVar = 0 . 10 , p < 0 . 001 , bootstrap test . Session 2: 1-CircVar = 0 . 09 , p < 0 . 001 , bootstrap test ) . The preferred orientation was very similar in both sessions ( 173° in session 1 , 168° in session 2 ) . Previous studies demonstrated that the gamma power of the LFP in V1 of monkeys is also tuned to orientation [33 , 34] . The moving grating stimuli elicited a clear peak in the LFP power spectrum of human V2/V3 , with a frequency between 50–60 Hz , and the power of this gamma oscillation at both electrodes was tuned for orientation ( E6: 1-CircVar = 0 . 11 , p < 0 . 001 , HWHM = 28° . E7: 1-CircVar = 0 . 12 , HWHM = 59° , p < 0 . 001 ) ( Fig 3C ) . We also observed a reliable influence of stimulus orientation on the peak frequency of the gamma oscillation ( determined by fitting a Gaussian function to the fractional increase in power per frequency bin; Fig 3D , Materials and Methods ) . Stimulus orientations eliciting strong gamma power caused oscillations with a higher frequency . Although the difference between the highest and lowest frequencies was only 4–5 Hz , the effect was very reliable ( bootstrap test: p < 0 . 001 for both electrodes; Fig 3E ) . Thus , the power and frequency of the gamma oscillations were well tuned for orientation . The tuning of gamma matched the tuning of spikes at E7 because the preferred MUA orientation was 171° , the preferred orientation of gamma power was 158° , and gratings with an orientation of 167° evoked oscillations with the highest frequency . Measurement of the spatial frequency tuning indicated that neurons at both E6 and E7 exhibited low-pass spatial frequency tuning ( Fig 3F ) . The strongest responses were evoked by gratings of 0 . 5 cyc . degree-1 , which was the lowest spatial frequency tested by us , so that we cannot exclude band-pass spatial frequency tuning , had the stimulus set contained even lower spatial frequencies . The spatial-frequency tuning of thresholded multi-unit activity ( MUAt ) was similar ( S2C Fig ) . The gamma power of the LFP also varied with spatial frequency . Tuning curves obtained from gamma power in the 30–100 Hz range showed a moderate correspondence to those obtained with MUA with stronger power at low spatial frequencies ( Fig 3G ) . The shape of the gamma peak in the power spectrum varied with spatial frequency ( Fig 3H ) . The grating of 0 . 5 cyc . deg-1 produced a broad gamma peak at high frequencies ( 60–90 Hz ) whereas higher spatial frequencies ( 1 , 2 cyc . deg-1 ) produced tighter peaks at lower frequencies ( 40–60 Hz ) , and the highest spatial frequencies tested ( 4 , 8 cyc . deg-1 ) did not produce any detectable gamma peak . We next examined tuning to contrast with the drifting gratings . As expected , the MUA response increased with contrast and saturated at high contrasts ( Fig 4A and 4B ) ( we obtained similar results with MUAt; S2D Fig ) . We fit a Naka-Rushton function to the contrast-response data to estimate the point at which MUA was 50% of its maximum ( C50 ) and obtained a C50 of 7 . 3% for E6 and 4 . 2% for E7 . These values are relatively low , yet within the range observed in the early visual cortex of monkeys [3 , 31 , 32] . The latency of the MUA response evoked by high-contrast gratings was shorter than that evoked by low-contrast gratings ( Fig 4A and 4B ) , as has also been observed in monkey V1 [35 , 36] . We also examined the influence of contrast on the oscillations in the LFP . The amplitude of the gamma rhythm in the LFP of both E6 and E7 ( Fig 4C ) became stronger with increasing contrast ( Fig 4C and 4D ) , and the peak frequency also became higher , in particular for the grating with 100% contrast , just as been observed in the visual cortex of monkeys [37–40] . At the same time , the grating stimuli with contrasts higher than 2% decreased LFP power at lower frequencies ( 10–25 Hz , the alpha to beta range ) ( Fig 4C and 4D ) , confirming results in monkey visual cortex [41 , 42] . Neurons in early visual cortex are tuned to the size of a visual stimulus . When the size of a stimulus is increased , neural responses are initially enhanced up to an optimum size . After this point , the response of many neurons decreases for larger sizes and it then reaches a stable level for very large sizes [43 , 44] . This response profile is well modeled by a ratio-of-Gaussians model with an excitatory receptive-field center and a larger , superimposed suppressive surround [44 , 45] . The responses of neurons at electrodes E6 and E7 were also tuned for the size of drifting gratings ( Fig 5A ) and their size-tuning profiles were well fit by the ratio-of-Gaussians model with optimum sizes of 3 . 7° ( E6 ) and 3 . 3° ( E7 ) ( Fig 5B; the MUAt results were similar , see S2E Fig ) . These size-tuning profiles are in accordance with the size tuning of neurons in area V2 of the macaque monkey [46] and there is , to our knowledge , no data available from V3 . We also computed the surround index ( SI ) to quantify the strength of the suppressive surround . We obtained SIs of 0 . 24 for E6 and 0 . 31 for E7 , values which are also well within the range observed in V2 of monkeys [46] . We next examined the influence of stimulus size on the gamma oscillations of the LFP , because previous studies in monkeys demonstrated an influence of this factor on the strength and frequency of gamma oscillations [47 , 48] . The size of the grating stimulus had a strong influence on the LFP oscillations . The largest grating ( 10 degree diameter ) suppressed the alpha oscillations that characterized the pre-stimulus epoch and replaced them with strong gamma oscillations , an effect visible in individual trials ( Fig 5C ) . We fitted a Gaussian to the peak in the amplitude spectrum to examine the influence of stimulus size on the amplitude and frequency of the oscillation . Larger stimuli increased the amplitude of the gamma oscillations with a strong linear relationship between gamma amplitude and stimulus size ( r2 = 0 . 96 , p < 0 . 001 ) and they decreased the gamma frequency ( Fig 5D , 5E and 5G ) . Thus , the influence of stimulus size on MUA ( Fig 5B ) differed from the effect on LFP gamma power ( Fig 5D and 5G ) , and accordingly , the linear correlation between LFP gamma amplitude and the MUA response was not significant ( r2 = 0 . 06 , p = 0 . 49 , Fig 5F ) . These results are in line with studies in area V1 of monkeys , which demonstrated that gamma power increases but spiking activity decreases when a grating stimulus encroaches into the suppressive surround of the receptive field [47 , 48] . The size tuning results illustrate that the responses of cells in the early visual cortex are modulated by contextual stimuli placed outside the RF [49] , because surround stimuli suppress responses to a stimulus placed in the center of the RF . Previous studies revealed that surround suppression depends on the relative orientation of the center and surround; surround stimuli which share the same orientation as the center produce more suppression than stimuli with an orthogonal orientation [50–52] . This effect , known as orientation-tuned surround suppression ( OTSS ) , is thought to enhance the representation of potential objects in the visual scene as it increases responses in regions with orientation contrast . Indeed , the perceived structure of the visual scene modulates neuronal activity in early visual cortex . Spiking activity elicited by figural regions is stronger than that elicited by background regions [4] . This figure-ground modulation ( FGM ) is delayed relative to the onset of the visual response and is thought to be due to feedback from higher visual areas [53 , 54] . We studied contextual modulation in V2/V3 using a paradigm in which OTSS and FGM could be measured using the same set of stimuli . We presented stationary gratings with the same basic properties as those used in the previous section on size-tuning . Every stimulus contained a central grating of 6° diameter , which was either presented in isolation or surrounded by a full-screen grating ( Fig 6A ) . The surround grating had the same orientation as the center ( Iso , Iso90 conditions ) or an orthogonal orientation ( Cross condition ) . In the Iso90 condition , the phase of the surround grating was shifted by 90° to create a region that perceptually segregated from the background . To measure the influence of OTSS on neuronal activity , we compared the Cross to the Iso90 condition . Both stimuli induce a figure-ground percept but orientation contrast is only present in the Cross condition . To measure FGM we compared the Iso90 condition with the Iso condition . These two stimuli do not have orientation contrast , but figure-ground segmentation only occurs in the Iso90 condition . The activity of neurons at E6 and E7 was strongly modulated by the surround gratings . The initial response was the same for all conditions , but after approximately 60 ms the response became suppressed in the Iso and Iso90 conditions relative to the Cross and Center only conditions ( Fig 6B ) . Thus , in this early phase the suppression depended purely on the orientation contrast between center and surround . After approximately 100 ms , the response in the Iso90 condition became enhanced relative to the Iso condition , and it stayed at an elevated level throughout the remainder of the trial . In this late phase , the neural response therefore appeared to reflect the perceptual segregation of the center region from the background . For statistical analysis , we calculated OTSS and FGM in an early time window ( 50–100 ms ) and a late time window ( 100–500 ms ) ( Fig 6C ) . In the early window , OTSS , calculated as the average difference between Cross and Iso90 , was significantly greater than zero for both E6 and E7 ( t tests , both ps < 0 . 01 ) but FGM was not ( both ps > 0 . 1 ) . In the late time window we observed the exact opposite pattern . Now FGM was significant ( both ps < 0 . 001 ) whereas OTSS was not ( both ps > 0 . 1 ) . We estimated the latency of both effects by fitting a Gaussian function to the time course of the modulation ( Fig 6D ) , averaging across the activity at E6 and E7 ( which were very similar ) ( Fig 6E ) . The latency of OTSS , defined as the moment where the Gaussian function reached 50% of its maximum , was 55 ms , which was approximately 10 ms after the latency of the visual response in the center-only condition ( dashed line in Fig 6E ) . FGM occurred at 98 ms , which was 43 ms after OTSS . These latencies are remarkably similar to those reported in monkey V1 for OTSS ( e . g . , ~60 ms in Bair et al . [55] ) and FGM ( ~100 ms in Self et al . , Lamme et al . , and Poort et al . [54 , 56 , 57] ) . In our main experiments we used a 6° diameter center stimulus so that the boundary between the center and surround fell outside the RF , but we obtained virtually identical results with a smaller ( 4° ) center stimulus that encroached into the RF ( S5 Fig ) . Apparently , the presence of the boundary in the RF was not an important factor for OTSS and FGM . These results , taken together , therefore suggest that the processes responsible for OTSS and FGM in the early visual cortex of humans are similar to those in monkeys . Previous studies demonstrated that the activity of neurons in early visual cortex of monkeys does not only depend on the stimulus in the receptive field and the surround but also on behavioral relevance . Specifically , attention shifts towards a stimulus in the receptive field enhance neuronal activity [5 , 58 , 59] . fMRI , EEG , and ECoG studies in humans support these findings [60–62] , but some of them reported that the influence of attention is weak [17] and these methods could not resolve whether the attentional effect causes subthreshold membrane potential changes or if it also leads to changes in spiking activity . To test the influence of attention on the spiking activity in human V2/V3 we used a curve-tracing task ( Fig 7A ) , in which the patient mentally traced a target curve that started at the fixation point to locate a larger red circle at the other end of this curve . Previous psychophysical studies demonstrated that human subjects gradually spread object-based attention over such a target curve [63] . Studies in monkeys revealed a neuronal correlate of the spread of object-based attention because traced curves elicit stronger spiking activity in early visual cortex than curves that are ignored [5 , 64] . When the stimulus appeared on the screen , the patient traced the target curve connected to the fixation point while maintaining gaze within a 2° radius window centered on the fixation point . After a delay of 500 ms , the fixation point changed color from red to green , and this was the cue to make an eye movement to a window centered on the larger circle at the end of this curve ( Fig 7A ) . The total number of stimuli was nine , because each of the three curves could be connected to the fixation point with equal probability , and we presented a total of three configurations that varied in the number of crossings between the curves . We first gave the patient a few minutes to practice before we started the recordings . After these practice trials , the patient performed the task with an accuracy of 90%; on 4% of trials she selected the wrong eye-movement target , and on the remaining 6% of trials the saccade did not enter any of the eye movement windows . We examined spiking activity of neurons at E6 and E7 on correct trials during the fixation period when the patient maintained gaze on the fixation point ( Fig 7B ) and excluded trials in which the patient made microsaccades during this period ( maximum velocity threshold of 10° . s-1 maintained for at least 10 ms; see S6 Fig for details of the eye movement analysis ) . Because the total number of trials was relatively small ( n = 135 correct trials ) we averaged the neuronal responses on correct trials across the three stimulus configurations . The initial , visually driven response did not distinguish between the target and distractor curves ( time window , 0–150 ms , t test , E6: p = 0 . 41 , E7: p = 0 . 32 ) . However , in a later response period ( 200–500 ms ) , the MUA responses evoked by the target curve were significantly stronger than those evoked by the distractor curves ( t test , E6: p = 0 . 04 , E7: p = 0 . 01 ) . The activity elicited by the distractor curve did not depend on whether the target curve was adjacent to the receptive field or farther away ( t test , E6: p = 0 . 84 , E7: p = 0 . 61 , Fig 7D ) . We also examined whether small changes in eye position or speed within the fixation window could account for the difference in activity evoked by the target and distractor curves , but the mean eye position and eye velocity were similar ( U-tests: x-position , p = 0 . 33; y-position , p = 0 . 16; velocity , p = 0 . 33 ) ( S6 Fig ) . Thus , visual attention modulates spiking activity in human area V2/V3 .
One of our aims was to compare the spiking activity in human and macaque early visual cortex . In making this comparison , we were faced with the difficulty that most recordings in macaque early visual cortex have been performed in V1 , some in V2 , and only relatively few in V3 . Furthermore , the RFs measured by us had an eccentricity of 10° , where recordings in macaques are rarely made . The many correspondences are nevertheless noteworthy . The size of the RFs were similar to those reported for MUA in V3 [23–25] . Contrast semi-saturation values ( C50 ) were 4%–8% , which is within the range expected in V1 [3] , V2 [31] , and V3 [32] . Furthermore , the strength of surround suppression was compatible with that in macaque V2 and the preferred stimulus size was similar [46] . Taken together , these results suggest that the tuning of cells in the early visual cortex of humans is well approximated by that in macaques . We observed large visually driven increases in gamma-band power of the LFP that were visible in individual trials ( Fig 5C ) . Brief flashes that we used to map the RFs elicited broad-band gamma oscillations from a region that had a similar size as the MUA RF . Thus , the gamma band of the LFP can be used to estimate the RF location and size if spiking data is not available . Moving grating stimuli elicited gamma oscillations in the LFP with power that was restricted to a narrower frequency band than the flashing stimulus [65] . The power and frequency of these narrow-band oscillations depended on the orientation , contrast , and size of the grating . The optimal orientation elicited oscillations that had more power and a higher frequency than stimuli with suboptimal orientations , in accordance with studies in cat [66] and macaque V1 [33 , 48 , 67] . High-contrast stimuli produced large increases in gamma power [37] , and the very highest contrasts led to oscillations at higher frequencies , as has been reported in macaque V1 [37 , 39 , 40] and V2 [40] . When we varied the orientation or contrast of the stimulus , increases in MUA were associated with increases in gamma frequency and power . However , the tuning of MUA and gamma to stimulus size differed . Larger grating stimuli suppressed the MUA and caused gamma oscillations with a lower frequency , but the gamma power increased with stimulus size , in accordance with results from macaque V1 [46] . The underlying mechanisms leading to this dissociation remain unknown , but it has been suggested that the strength of gamma oscillations depends critically on the balance between excitation and inhibition in cortex [47 , 48 , 68–70] . Thus , the tuning of the gamma-band oscillations in humans closely resembles that in the macaque . This generalization from the macaque to the human is important because insights into the relationship between gamma and spiking in macaques may facilitate the interpretation of ECoG , MEG , and EEG data in humans and its relation with the tuning of cells . It should be noted , however , that grating stimuli evoke strong gamma oscillations , whereas complex , more naturalistic stimuli evoke weaker gamma oscillations . Thus , the relationship between gamma and the firing of neurons is stimulus dependent [65] . We observed that responses in human V2/V3 exhibit two well-known contextual effects: orientation-tuned surround-suppression and figure-ground modulation ( Fig 6 ) . Image regions with orientation contrast and regions that belonged to figures elicited stronger spiking responses than image regions with a uniform orientation , even if the stimulus in the RF was held constant . The time courses of these contextual effects were very similar to those reported in macaque V1 [55 , 56] . This result suggests that V2/V3 in humans is targeted by feedback from higher visual areas that carry contextual information from larger regions of the visual scene , just like V1 in the monkey [54 , 71–73] . Our results demonstrate that cells in early visual cortex in humans do not only extract local feature information but that they may also play an active role in perceptual processes such as scene segmentation and figure-ground assignment , and they thereby confirm previous EEG and fMRI studies [74 , 75] . The attentional modulation that we observed in a curve-tracing task ( Fig 7 ) presents another striking similarity with previous results in the visual cortex of monkeys . Just as in monkey V1 [5] , the initial response of the neurons in human V2/V3 did not discriminate between a target and a distractor curve , but after a delay , the response elicited by the target curve was enhanced . Psychophysical studies [63 , 76] demonstrated that human observers spread object-based attention across the target curve and the modulation of neuronal responses in V2/V3 can therefore be explained by the spread of attention [77] . One possible concern in the interpretation of these results is that the data came from an epileptic patient who reported visual disturbances prior to the onset of the seizure . However it is unlikely that the cortical region we were recording from was structurally abnormal , for a number of reasons . Firstly , although the patient reported that a feeling of visual movement or “fluttering” sometimes preceded the seizures , they were not triggered by visual stimuli , and her visual performance and acuity were normal . Secondly , we observed no lesions or abnormalities in the anatomical MRI images or functional maps of her visual cortex . Thirdly , the clinical analysis of the intracranial EEG recordings suggested an onset zone outside of visual cortex close to the temporo-parietal junction . Lastly , the tuning of the neurons to visual stimuli corresponded well to those observed in the monkey , which seems unlikely for grossly abnormal or damaged cortical circuitry . We recorded from only two microwires in V2/V3 , and we do not foresee opportunities to increase the size of our sample in the near future , given the extreme rarity with which depth electrodes are placed in early visual cortex . At the same time , we were able to carry out many tests for the neurons at these two recording electrodes , because the recordings were stable across a number of days . Another limitation is that we recorded MUA rather than the activity of single neurons . The tuning of MUA reflects the average across a population of neurons in the vicinity of the electrodes in V2/V3 . On the one hand , tuning to stimulus direction and orientation was presumably less sharp than that of single neurons , because MUA implicitly averages activity across a number of neurons with different tuning curves . This limitation is less severe for tuning to size or contrast , because pooling across neurons does provide insight into the average dependence of firing rates within a cortical region on these factors . This also holds true for contextual and attentional effects . Contextual effects in early visual cortex are highly consistent across neurons [4] , and attention usually increases neuronal firing rates so that averaging across cells provides insight into the magnitude of the average effect [78] . The data presented here represents the first quantitative study of spiking activity in human early visual cortex . Using several different visual paradigms , we found remarkable similarities between the spiking of human visual neurons and those in macaque visual cortex , both in the basic tuning properties as well as in contextual modulation and attentional paradigms . These results confirm that the macaque monkey is an excellent model system for studying the properties of cells in the early visual system of humans .
The study was approved by the medical ethics board of the Free University Medical Center ( protocol 2009/194 ) . The patient gave written informed consent to participate in the study . She was a 35-y-old female with a normal IQ and no reported visual deficits . The patient had late-onset epilepsy without any relevant antecedents . Her semeiology ( seizure symptoms ) strongly suggested the involvement of posterior temporal cortex , or the visual cortex without clear lateralization . She was implanted with 15 depth electrodes ( AdTech ) under general anesthesia using frameless stereotaxy . The electrodes were inserted through a guide-tube under the guidance of an online stereotactic positioning system . Five electrodes were placed in the right hemisphere , and nine electrodes in the left hemisphere ( to determine lateralization of the seizure onset ) . One electrode in visual cortex was selected to contain a microwire bundle . Recordings from the microwires began 2 d after the surgery and proceeded for a further 7 d . The impedance of the side-wire electrodes was measured post-explantation as 158 KΩ ( E6 ) and 115 KΩ ( E7 ) . Signals from the microwires were amplified with respect to a skull-screw ground using a unity gain HS-9 head-stage amplifier ( NeuraLynx ) . We digitized and sampled the signal at 32 . 5 kHz before storing it for later analysis . The raw signals from E6 and E7 were re-referenced to the average of the ( nonspiking ) electrodes E0–E4 ( E5 was excluded from the average due to high noise levels ) . From the re-referenced signal we created three signals: the local field potential ( LFP ) , the envelope of multi-unit activity ( MUA ) and the thresholded multi-unit activity ( MUAt ) . The LFP was created by first down-sampling to 930 Hz , then band-pass filtering the resulting signal between 1 Hz and 200 Hz using a second order , zero-phase Butterworth filter . Line-noise was removed by fitting a 50 Hz sine-wave to each individual trial , then subtracting it . S1 Fig outlines our procedure for generating MUA and MUAt . Briefly , we measured MUA by band-passing the raw signal between 500 Hz and 5 kHz to isolate high-frequency ( spiking ) activity . This filtered signal was rectified ( negative becomes positive ) , down-sampled to 930 Hz and low-pass filtered ( <200 Hz ) to measure the envelope of the spiking activity . The MUA provides an instantaneous , threshold-free estimate of spiking activity in the vicinity of the microwire , and it is a good measure of the average single-cell activity [79] within 150 μm of the electrode tip [54] . We also generated MUAt by thresholding the band-passed signal ( see S1 Fig for details ) . We generated visual stimuli in MATLAB using the COGENT graphics toolbox developed by John Romaya at the LON at the Wellcome Department of Imaging Neuroscience and custom scripts . Stimuli were presented at 60 Hz on a laptop LCD screen ( 26 cm width ) located at a distance of 66 cm from the patient . The LCD screen had a mean luminance of 40 cd . m-2 and this value was used as the background gray in all experiments except where indicated . The patient viewed the screen in a dimly lit room with her head on a chin rest , and we measured her eye movements with an Eyelink T1000 system sampling at 1 , 000 Hz . In all sessions , the patient began each trial by fixating on a circular fixation pattern presented towards the top-left corner of the laptop screen ( so that we could place stimuli in the neurons’ RFs ) . She maintained fixation within a 2° radius circular window during presentation of the stimuli; otherwise , the trial was aborted . Analysis of her eye position showed that she actually maintained fixation in a much smaller area than this ( S6 Fig ) . For the experiments measuring contextual and attentional modulation , we discarded trials in which the eye position was further than 1° from the center of fixation ( 10%–15% of trials ) . We measured RFs by flashing a small black-and-white checkerboard pattern ( 1° x 1° size , check-size 0 . 33° ) at every point of an 11° x 11° grid . RF tuning properties were measured using drifting sine-wave gratings placed at a location that activated neurons at both E6 and E7 ( 10 . 3° eccentricity , -14° angle from horizontal meridian ) . As standard stimuli , we used gratings with a Michelson contrast of 80% that were 10° in diameter with a spatial frequency of 1 cycle/degree and drifted with a temporal frequency of 3 Hz . We presented orientations of 45° or 135° , except in the orientation/direction tuning session , which included 24 directions ranging from 0 to 345° , in steps of 15° . For the other tuning sessions , we varied one parameter while holding the others constant . Each session contained ten repeats of every stimulus parameter . We completed two sessions for the orientation-tuning and contrast-tuning experiments and one session for the spatial frequency and size tuning experiments . In the contextual modulation experiments , the gratings had the same properties as those described above except that they were stationary . The phase of the central grating was randomly chosen on each trial from a uniform distribution ranging from -π to π . We used two different , orthogonal orientations for the central grating , 60° and 150° . We presented an equal number of trials with the two orientations so that the average stimulation of the RF was identical for all conditions . The stimulus duration was 500 ms with an inter-trial interval of 500 ms . In the curve-tracing experiments the patient began each trial by directing gaze to the fixation point . After 300 ms the curves and targets appeared . The curves had a width of 5 pixels ( 0 . 1° ) . The targets were red , 3° in diameter and were presented at -10° , -30° , and -50° from the horizontal meridian at 13° eccentricity . We presented a total of nine stimuli in a pseudo-random sequence ( Fig 7A ) . The patient had to maintain fixation for a further 500 ms , and then the fixation point became green , cueing the patient to make a saccade to the larger circle at the end of the target curve . For the MUA analysis , we first subtracted the average background activity across all trials ( -0 . 2 s to 0 s before stimulus onset ) . The data from the RF mapping experiment ( shown in Fig 2 ) were normalized by dividing by the average activity in a window from 0 . 05–0 . 3 s from the position that gave the strongest response . All other MUA data were normalized by dividing by the peak response in a time window between 0 . 03 and 0 . 15 s . In the RF tuning experiments we normalized to the maximum activity across conditions , in the contextual modulation experiment to the maximum response in the center-only condition , and in the curve-tracing experiments to the maximum response averaged across all trials . For MUAt analysis we first created histograms by binning the spike-times from each individual trial into bins of 1 . 1 ms duration . We then convolved the resulting spike-trains with a Gaussian density function to create spike-density functions . See S1 Fig for further details . For LFP analysis of the RF tuning data , we computed the Fourier transform of the LFP in two windows , a pre-trial baseline ( -0 . 3 to 0 s before stimulus onset ) and during the stationary period of the response ( 0 . 15–0 . 45 s after stimulus onset ) . We applied a Hann window before computing the Fourier transform to reduce edge artifacts . We estimated the power in each frequency bin as the squared magnitude of the signal and computed the relative change in power by dividing the post-stimulus power by the baseline power spectrum . The patient was scanned three months after the explantation of the electrodes on a 3T MRI scanner ( Prisma Fit; Siemens Medical Systems , Erlangen , Germany ) equipped with a 64-channel head coil . We collected anatomical T1-weighted images using a magnetization-prepared rapid-acquisition gradient echo ( MPRAGE ) pulse sequence ( 192 sagittal slices; Repetition Time [TR] = 2 , 250 ms; Echo Time [TE] = 2 . 21 ms; Flip Angle [FA] = 9°; Field of View [FoV] = 256 × 256 mm2; 1 mm isotropic resolution; GRAPPA = 2 ) . Functional data were acquired using a gradient-echo echo-planar imaging sequence ( 30 transversal slices; TR = 1 , 000 ms; TE = 30 ms; FA = 60°; FoV = 216 × 216 mm2; 2 mm isotropic resolution; no slice gap; MultiBand factor = 2; GRAPPA = 2 ) . We analyzed the imaging data with BrainVoyager QX ( v2 . 8 . 4; Brain Innovation , Maastricht , the Netherlands ) . Anatomical data underwent brain extraction , followed by inhomogeneity correction and semi-automatic segmentation of the gray-white matter boundary for mesh reconstruction of the cortical surface . Preprocessing of the functional data included slice scan time correction , ( rigid body ) motion correction , and temporal high-pass filtering ( up to two cycles per run ) . To estimate the location of the electrode contacts , the post-implantation CT scan was aligned to the anatomical MR data , and the clearly visible macrocontacts were segmented for visualization in combination with the subject’s cortex mesh reconstruction . Functional localization relied on delineation of early visual areas based on three population Receptive Field ( pRF ) mapping runs ( each lasting ~4 min ) [80] . In every run , a bar stimulus ( projected onto a back-projection screen visible to the subject via a mirror mounted onto the head coil ) was semi-randomly presented twice for 2 s at 12 different locations and four orientations ( spanning a total of 20° by 20° ) using the StimulGL presentation software [81] . For the analysis of the receptive-field location , we took the average activity ( either MUA or gamma power ) in response to each flash location in a time window between 50–300 ms ( MUA ) or 100–250 ms ( gamma power ) . The responses , R ( x , y ) , at each x and y location were normalized by dividing by the maximum response across all locations , and we fit a 2D , elliptical Gaussian using nonlinear multidimensional minimization: R ( x , y ) = e-a ( x-rfx ) 2+2b ( x-rfx ) ( y-rfy ) +c ( y-rfy ) 2 where a = cos2θ2σx2+sin2θ2σy2 , b = -sin2θ4σx2+sin2θ4σy2 and c = sin2θ2σx2+cos2θ2σy2 . The center of the Gaussian ( rfx , rfy ) , the standard deviations in the x and y directions , σx and σy and the orientation of the Gaussian θ are the free parameters . Orientation tuning width was assessed by fitting a circular Gaussian using nonlinear least-squares fitting in MATLAB as follows: R ( θ ) = C+ Rp∙exp ( -angori ( θ-θpref ) 22σ2 ) , Where R ( θ ) is the response between 0–0 . 5 s for orientation θ , C is the offset , Rp is the response to the preferred orientation , angori ( x ) = min ( abs ( x ) , abs ( x-180 ) , abs ( x+180 ) ) wraps angular differences to the interval 0° to 90° , and σ is the standard deviation of the Gaussian . Tuning width was measured as the orientation difference between the peak of the response and the midpoint between the peak and the minimum value ( half width at half maximum , HWHM ) . Tuning strength was assessed using 1-CircVar [82] which has been shown to be a more robust estimator of tuning strength than orientation indices [30] . It was calculated as follows: 1-CircVar = ∑kR ( θk ) e ( 2iθk ) ∑kR ( θk ) where R ( θk ) is the response to the orientation θk ( in radians ) . The significance of the tuning strength was assessed by bootstrapping . For each electrode , we randomly resampled with replacement an equal number of trials as in the original dataset but shuffled the responses across orientations and calculated 1-CircVar 10 , 000 times to create the expected null distribution of the 1-CircVar statistic from which we derived the p-value . We fitted a Naka-Rushton equation to the contrast response curves in a time window from 0 to 0 . 5 s: R ( x ) = bxNC50N+ xN where R ( x ) is the average response at a given contrast value , x , b controls the steepness of the curve , and C50 is the point at which the curve reaches 50% of its maximum , which is a measure of contrast sensitivity [3] . To measure size-tuning curves , we fit the average response between 0 and 0 . 5 s after stimulus onset with a ratio-of-Gaussians model ( ROG ) , which provides a good fit to size-tuning curves [44 , 45]: R ( x ) = Geerf ( xWe ) 21+Gierf ( xWi ) 2 Ge , Gi , Wi , and We are the gains and widths of the excitatory center and inhibitory surround , respectively; x is the size of the grating in degrees; R ( x ) is the response; and erf is the error function . The surround suppression index ( SI ) was calculated as: SI = Rmax-RsuppRmax where Rmax was the maximum of the modeled response and Rsupp was the response of the model to the largest size . To estimate the amplitude , peak frequency , and width of the gamma peak , we fit a Gaussian function to the relative change in gamma power between 30–100 Hz using nonlinear multidimensional minimalization with the Nelder-Mead algorithm in MATLAB: P ( f ) = Ge-12 ( f-μσ ) 22πσ+b , where P ( f ) is the power at frequency f , G determines the amplitude , b is an offset , μ is the peak frequency , and σ is the standard deviation . | Our knowledge of the function of the early visual cortex is based largely on recordings of spiking activity from neurons in animal models , in particular the macaque monkey . Indirect measurements of neuronal activity in the human visual cortex have suggested many similarities with the macaque visual cortex , but to date there have been no quantitative analyses of spiking data in the human early visual cortex . In this paper , we report spiking data recorded from the early visual cortex of a patient who was implanted with depth electrodes as part of her treatment for epilepsy . We were able to verify that human visual neurons have response properties similar to macaque neurons , including the size of their receptive fields and the presence of orientation tuning . We also found that the responses of human visual neurons are modulated by the visual context and by shifts of attention in a virtually identical manner to neurons in the macaque . This study , therefore , shows that the macaque visual system provides an excellent model for human visual cortical processing . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"engineering",
"and",
"technology",
"signal",
"processing",
"membrane",
"potential",
"brain",
"vertebrates",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"animals",
"mammals",
"signal",
"filter... | 2016 | The Effects of Context and Attention on Spiking Activity in Human Early Visual Cortex |
The molecular details underlying the time-dependent assembly of protein complexes in cellular networks , such as those that occur during differentiation , are largely unexplored . Focusing on the calcium-induced differentiation of primary human keratinocytes as a model system for a major cellular reorganization process , we look at the expression of genes whose products are involved in manually-annotated protein complexes . Clustering analyses revealed only moderate co-expression of functionally related proteins during differentiation . However , when we looked at protein complexes , we found that the majority ( 55% ) are composed of non-dynamic and dynamic gene products ( ‘di-chromatic’ ) , 19% are non-dynamic , and 26% only dynamic . Considering three-dimensional protein structures to predict steric interactions , we found that proteins encoded by dynamic genes frequently interact with a common non-dynamic protein in a mutually exclusive fashion . This suggests that during differentiation , complex assemblies may also change through variation in the abundance of proteins that compete for binding to common proteins as found in some cases for paralogous proteins . Considering the example of the TNF-α/NFκB signaling complex , we suggest that the same core complex can guide signals into diverse context-specific outputs by addition of time specific expressed subunits , while keeping other cellular functions constant . Thus , our analysis provides evidence that complex assembly with stable core components and competition could contribute to cell differentiation .
A key question in cellular network biology is how protein complexes assemble and disassemble in a time-dependent manner . Coordinated changes in the transcriptome and proteome occur during cellular differentiation [1–3] , during cell reprogramming [4] , or after growth factor stimulation [5] to name a few examples . Recent work in yeast has predicted that complexes change in composition during the cell cycle , and that complexes consist of both constitutive but non-dynamically expressed , and dynamically expressed subunits , leading to the proposal of ‘just-in-time assembly’ of complexes [6] . Consistent with this concept , relating expression data in different human cell types and tissues to protein complexes showed that non-dynamically expressed proteins extensively interact with tissue-specific expressed proteins , suggesting a tight interplay between core and tissue-specific proteins [7] . However , the molecular details underlying the assembly of complexes ( ‘complex assembly motifs’ ) are largely unexplored . This includes the definition of complexes themselves , e . g . as molecular machines ( stably associated complexes ) or pleiomorphic ensembles ( complexes that assemble on demand ) [8] . For example , what is the proportion of complexes that are permanently assembled , changed during different cellular conditions , or contain both non-dynamic and dynamic subunits ( ‘di-chromatic’ complexes ) ? Are subunits replaced at structurally overlapping or compatible surfaces of proteins ? What is the role of evolutionarily-related paralogs ? Complementing protein interaction networks with three-dimensional structural information for binding interfaces has provided an improved functional understanding of cellular protein networks [9–15] . A recent study combining structural modeling with network analyses has revealed two types of interactions with a common hub protein: mutually exclusive interactions through a single interface ( also called ‘XOR’ ) and compatible interactions through multiple interfaces ( also called ‘AND’ ) [10] . This study has also shown that hub proteins characterized by single interfaces evolve faster and are enriched in paralogs [10] . In another study , modeling ErbB signaling through combining network and structural analyses , has suggested that competing protein interactions at single interface hubs produce variations in signaling responses [14] . Based on the above studies , we reasoned that the assembly of complexes where proteins compete for a common stable core could play a role in cell differentiation . To test this hypothesis and to define complex assembly motifs , we focused on the calcium-induced differentiation of primary human keratinocytes ( PHK ) as a model system for a large cellular reorganization process [16 , 17] . The epidermis of mammalian skin develops from a single layer of keratinocytes ( interfollicular basal stem cells ) into a multi-layered stratified epithelium . Keratinocyte differentiation is a well-suited model system , as differentiation of primary keratinocyte cells can be induced in vitro in cell culture by the addition of calcium [18] . In our previous work we quantified the transcriptome during differentiation , and we identified functionally important circadian oscillations [3] . Here , we performed a detailed analysis of all gene expression changes associated with keratinocyte differentiation followed during 45 hours and integrated this information with protein complexes to analyze their reorganization . We inferred that half of human protein complexes present during differentiation contain both non-dynamically and dynamically expressed subunits . Some di-chromatic complexes contain a stable core that associates with dynamic genes belonging both to similar clusters ( concerted gene expression changes ) and different clusters ( opposing gene expression changes ) . In many cases , di-chromatic complexes with genes exhibiting opposing expression changes belong to complexes known to be involved in general or keratinocyte-specific differentiation processes and pathways , such as EGF/TGF-α signaling , TNF-α/NFκB signaling , Notch/γ-secretase , ubiquitination , cell cycle arrest , and chromatin remodeling complexes . Using three-dimensional structural modeling , we predicted physical interfaces and distinguished between mutually exclusive ( XOR ) and compatible ( AND ) interactions [13 , 19] . We found that dynamic proteins binding to a common non-dynamic protein are enriched for mutually exclusive interactions , suggesting that changes in complex assemblies can occur through variation in the abundance of proteins that compete for binding to XOR nodes . In addition , compensatory expression changes of paralogs suggest that these proteins—while keeping a constant essential function for cell viability—have differential functionalities , which serve a specific role for cell type-specific functions . Altogether our analysis highlights the importance of understanding the assembly of complexes and taking 3D structural information into consideration , rather than elucidating networks of individual proteins .
Differentiation was initiated in human primary keratinocyte stem cells by the addition of CaCl2 [3 , 18] . Cells were harvested over 45 hours at 5-hour intervals with three biological replicates . At each time point cells were lysed , mRNA was isolated , and expression profiles were measured using Agilent microarrays [3] ( Fig 1A and S1 Table ) . 21 , 113 probes mapping to 16 , 720 genes were detected as expressed . We defined genes that are changing ( ‘dynamic’ ) and those that are constantly expressed ( ‘non-dynamic’ ) using a 2-fold expression change cut-off and a Chi-squared test on time points of biological replicates ( see methods ) . As a more stringent criterion for classifying dynamic genes , we used a threshold of 4-fold , defining these genes as ‘super-dynamic’ ( S1 Table ) . In summary , out of the 16 , 720 expressed genes , 6 , 137 are ‘non-dynamic’ ( P> = 0 . 01 , Chi-squared test , and fold change <2 ) , 6 , 096 are ‘dynamic’ ( P<0 . 01 and fold change > = 2 ) , 1 , 317 of the dynamic genes are also ‘super-dynamic’ , ( P<0 . 01 and fold change > = 4 ) , and 4 , 487 genes do not fall into one of the above categories ( classified as ‘unresolved’ ) . In general , we used the super-dynamic genes for our analyses , unless otherwise stated . Classification of proteins encoded by the super-dynamic genes based on DAVID [20 , 21] , UniProt [22] , and manual literature searches ( e . g . [23] ) ( S1 Fig ) uncovered a large fraction of proteins involved in: i ) signaling ( transcription factors , and the adhesion , chemokine , calcium , immune/Toll , ubiquitin-like , apoptosis , Wnt , TGFβ , interferon , and Notch signaling pathways ( S2 Fig ) ( 32% ) ; ii ) housekeeping functions ( cytoskeleton , cell cycle , solute transport carriers , histone , and chaperone proteins , or formation of tight junctions ( S3 Fig ) ( 22% ) ; iii ) metabolic enzymes for lipid , amino acid , steroid , purine , and flavine interconversions and lipid binding ( S4 Fig ) ( 13% ) , iv ) proteins needed for the progressive steps towards the formation of the cornified envelope ( metallo and serine proteases ) , crosslinking enzymes ( e . g . transglutaminases ) and substrates ( e . g . loricrin , involucrin , and small proline rich proteins ) that provide structural stability and elasticity , keratins ( mechanical resistance ) , and lipid modifying enzymes ( water repellence ) ( S5 Fig ) ( 8% ) [23] . For 8% of the super-dynamic genes only one Pfam domain prediction can be assigned , while 17% of them have no known function or Pfam domain annotations . Thus , the process of keratinocyte differentiation is accompanied by concerted changes in metabolic , signaling , and housekeeping pathways . We also confirmed the induction of known markers for keratinocyte differentiation ( involucrin [IVL] , filagrin [FLG] , cystatin [CSTA , CSTB] ) ( S1 Table ) [23] . Involucrin expression was additionally confirmed on the protein level by immunostaining ( S6 Fig ) . We used K-means clustering to classify the temporal profiles of the 1 , 317 super-dynamic genes and identified eight optimized clusters ( Fig 1B ) . The clusters are arranged with opposing temporal profiles on the top and on the bottom: clusters 1 and 5 contain early highly transient expressed/repressed genes , clusters 2 and 6 contain early transient expressed/repressed genes , clusters 3 and 7 contain early highly sustained expressed/repressed genes , and clusters 4 and 8 contain early sustained expressed/repressed genes . Consistent with the function of the Ets transcription factor ELF3 in promoting differentiation of keratinocytes [24 , 25] , we see strong immediate and sustained expression of ELF3 ( cluster 3 ) , followed by delayed expression of genes regulated by ELF3 like KRT4 and TGFβ ( cluster 8 ) , and the envelope protein SPRR2A ( cluster 3 ) [26 , 27] . The clusters reveal a moderate stage-specific expression of functionally related protein classes and biological processes . For instance , proteins related to the formation of the cornified envelope are strongly induced at later time points ( clusters 3 , 4 , 5 , and 8 ) . Cluster 7 ( repressed genes ) is enriched for cell cycle-related proteins in the housekeeping category , as expected during the onset of differentiation [28] . This cluster has an antagonistic behavior when compared to the cell-cycle enriched cluster published in a cellular reprogramming study [4] . Signaling-related pathways in clusters 1 and 2 involve transiently expressed transcription factors , cytokines , and proteins involved in Wnt , adhesion , TGFβ , and TNF signaling ( S1 Table ) . Likewise , enzymes involved in lipid and amino acid metabolism are present in most clusters . The absence of a strong association between functional categories and clusters suggests a functional replacement within sub-categories , possibility brought about in some cases by dynamic rearrangements within protein complexes . During keratinocyte differentiation , a switch in the expression of paralogous gap junction subunits is needed for the changes in gap junction permeability required for epidermis formation [29] , which we see in our study ( Fig 2A ) . To see if this is a general case for all paralogous pairs , we annotated expressed genes in keratinocytes with homology information from EnsemblCompara [30] limiting our analysis to paralogous pairs in which both genes are dynamically expressed , thereby reducing our set from 11 , 582 to 2 , 260 paralogous pairs . Computing Pearson correlation metric for all pairs , we found 950 to have highly correlated ( r> = 0 . 6 , P<0 . 07 ) dynamic expression profiles during skin differentiation with a subset of 235 qualifying as super-dynamic , and 281 pairs to have highly anti-correlated dynamic expression ( r< = -0 . 6 , P<0 . 07 ) with 19 of these pairs being super-dynamic ( Fig 2B and S2 Table ) . When compared to super-dynamic and dynamic random pairs , we found that paralagous pairs are more likely to have correlated expression ( Wilcoxon rank sum test; P = 4 . 5e-37 , for super-dynamic S7A Fig; and P = 5 . 4e-61 for dynamic , S7B Fig ) . An example where gene duplicates can act together to bring about a functional change is the formation of cornified envelope ( e . g . the transglutaminase substrates , small proline-rich proteins ) ( Fig 2C and S8 Fig ) . This could suggest a requirement for increased gene dosage in particular specific stages of cell differentiation as the likely reason for the duplication of these genes [31–33] . We also found nineteen anti-correlated pairs where both proteins are super-dynamic ( Figs 2D–2F and S9 Fig ) . Examples include PLEKHA 6/7/4 ( Fig 2D; for PLEKHA-7 a functional role has been demonstrated in recruiting paracingulin to tight junctions [34] ) , WNT7A/7B/5A ( Fig 2E ) , NDRG1/2/4 ( Fig 2F; NDRG2 is expressed in response to TGFβ and inhibits proliferation [35] ) and CCNA/B proteins ( S9 Fig ) . One explanation for this observation might be that there is temporal sub-functionalization , with one paralog being expressed early , the other late during the course of differentiation as the paralogs have taken on different functions . Therefore , the expectation would be that anti-correlated paralogous pairs are more divergent in sequence level than correlated ones . Using Pfam domain annotations we compared correlated to anti-correlated dynamic paralogous pairs ( here we used dynamic pairs to increase the numbers for statistical analysis since not every protein has Pfam annotations ) and found that correlated paralogous pairs are more similar in their Pfam domain composition ( different domain , or domain missing ) than anti-correlated pairs ( 1 . 3 versus 1 . 7; Wilcoxon rank sum test; P = 0 . 009 ) ( Fig 2G and S10 Fig ) . Likewise , anti-correlated dynamic paralogous pairs have greater sequence divergence ( S11A and S11B Fig ) and a greater number of differences in amino acid sequence length ( S11C and S11D Fig ) . Furthermore , anti-correlated genes have a higher fraction of evolutionary old duplicated genes compared to correlated genes ( S12 Fig ) . In summary , dynamic paralogs are enriched in gene pairs that are correlated in expression . This suggests that a requirement for increased gene dosage is the likely reason for the duplication of these genes . Paralogous genes with anti-correlated expression changes have greater amino acid sequence and Pfam domain divergence . This implies that anti-correlated paralogs have taken on specialized roles during differentiation . The finding that divergent paralogous pairs tend to have anti-correlated expression patterns could suggest replacement of subunits in protein complexes to expand functional complexity . To see if this is a general feature of protein complexes during differentiation , we obtained a list of protein complexes from the CORUM database ( a resource of manually annotated protein complexes from mammalian organisms [36] ) , and mapped it onto the non-dynamically expressed , unresolved , or dynamically changing proteins ( Fig 3A and S13 Fig and S3 Table ) . After removing complexes containing unresolved genes , 19% of the complexes did not have significantly changing gene expression for any of their subunits during the keratinocyte differentiation process ( all non-dynamic ) , 26% of the complexes had dynamically changing expression for all subunits , and for half of the complexes ( 55% ) we find a mix of behaviors ( here called ‘di-chromatic’ ) ( Fig 3B ) . We represented the previously clustered 1 , 317 super-dynamic genes in the context of complexes to which they belong and also assigned dynamic genes to each one of the eight super-dynamic clusters through comparative correlation analysis with average expression profile of the aforesaid clusters ( Fig 4; S1 Network ) . We found that super-dynamic genes in the same complex or closely connected in the network often belong to different clusters ( 72% ) . However , this ratio decreases to 43% if clusters with similar behaviors ( 1 and 2 , 3 and 4 , 5 and 8 , 6 and 7 ) are combined . To further characterize the expression profiles of di-chromatic complexes , we classified the CORUM complexes into functional groups ( S3 Table ) . Then we represented the expression changes for functionally related complexes or complexes sharing components ( Fig 5 and S14–S16 Figs ) . Among many cellular processes known to be important for keratinocyte differentiation or general cell differentiation ( Fig 5A ) , we find examples of complexes with a stable component or core and a dynamic periphery . Examples in cell signaling include EGFR and TNF-α/NFκB pathways . EGF signaling is important for keratinocyte differentiation [37] and we find two complexes involved in its recycling and degradation . In both complexes , one component ( STAM2 and CBLB ) is constant and the other is super-dynamic ( RIN1 and SH3BKP1 , respectively ) , changing in opposite directions in the two complexes . RIN1 regulates EGFR degradation in cooperation with STAM [38] , and SH3BKP1 prevents epidermal growth factor receptor degradation by the interruption of c-Cbl-CIN85 complex [39] . Thus , the opposite behavior of these complexes will favor EGFR stabilization and consequently , EGF signaling and keratinocyte differentiation . In this respect we also find anti-correlated changes of the paralogous receptors EGFR and ErbB2 , which slow down EGFR recycling through heterodimer formation [40] . The TNF-α/NF-κB signaling pathway is required for normal epidermal development and homeostasis [41–43] . As the CORUM database is not complete [44] , we combined the CORUM complex information with a detailed structural and pathway analysis based on the literature [45 , 46] ( S15 Fig ) . We identified a core TNF receptors/scaffold complex ( TNFRSF1 , TRADD , RIPK1 , TRAF2 , IKBG1 , CHUK , CDC37 and HSP90AA1 ) associated with a MAPK pathway ( i . e . MAP2K7/ and MAPK8/9 ) which is composed of non-dynamic and unresolved genes , while the rest of the pathway in general is composed of dynamic components . This is an example of a stable core module associated with a dynamic peripheral module . Most of the super-dynamic genes are found mainly in the transiently expressed clusters 1 and 2 , suggesting relevant concerted changes during keratinocyte differentiation ( Fig 5B ) . MAP3K8 ( Cot ) is in cluster 4 ( early activated and sustained expression ) and can form larger complexes containing NFKB1 , thereby promoting signaling . Likewise , CFLAR is in the same cluster 4 and is an inhibitor of FADD thereby preventing apoptosis . NF-κB/RelA has been shown to control cell-cycle exit in keratinocytes [47] . NF-κB-independent signaling commencing at the level of the non-dynamic core connects to MAPKs ( such as NIK , NAK , TAK1 , and Cot ) and PKC isoforms [48] . Interestingly we see that TRAF1 expression is in the same cluster as the NF-κB components . It has been described that TRAF1 promoter has NF-κB binding sites and is strongly activated by TNF-α [49] . TRAF1 cannot bind to the TNF receptor directly , but is recruited though binding to TRAF2 [50] and overexpressing TRAF1 does not affect the interaction between TRAF2 and FADD [51] , suggesting a compatible complex formation . TRAF1 binding to TRAF2 results in blocking apoptosis [52] and analysis of TRAF1-/- mice suggests that TRAF1 inhibits TNF-α signaling [49] . Thus , TNF-α/NF-κB induced dynamic expression of TRAF1 creates a negative feedback loop that could mediate anti-apoptotic functions [52] and decrease NF-κB activation without affecting other possible constitutive function of TNF receptor . We propose that the stable core module connects to both a dynamic peripheral module important for cell cycle arrest during differentiation ( via NF-κB heterodimers and prevents apoptosis ( via TRAF1 ) , and several stable modules that function independent of NF-κB by binding to TRAF or the IKK complex and which should play a housekeeping function [48] ( see S15 Fig for a summary of different signaling functions ) . In the Notch/γ-secretase pathway [53] , the γ-secretase ( APH1B , promoting differentiation ) is up-regulated and delta like 1 ( DLL1 , blocking differentiation ) is down-regulated ( Fig 5B ) . Both proteins associate to the same stable core ( PSENEN and NCSTN ) , reinforcing a biological role for di-chromatic complexes . In the ubiquitination/degradation pathway , the CAND1 assembly factor of the E3 ubiquitin ligase complex [54] is up-regulated transiently and the S-phase kinase-associated protein 2 ( SKP2 , promoting cell cycle ) is down-regulated as the target of the E3 ubiquitin ligase complex ( S14A Fig ) . Skp2–Skp1 abrogates the inhibitory influence of CAND1 on the neddylation of Cul1 by promoting the dissociation of the cullin–CAND1 complex , whereas substrate , together with substrate-presenting components , prevents the action of CSN to deneddylate cullin [55] . It has been described that high levels of Skp2 are needed for proliferation in stratified epithelia [56] . In this respect it is noteworthy that CAND1 also causes elevation of p27 , which has been demonstrated to be important during pre-adipocyte differentiation [57] . Finally , two complexes with interesting di-chromatic behavior are related to chromatin remodeling and known to be important for regulating gene expression changes during keratinocyte differentiation [58 , 59] . Emerin is a nuclear membrane protein , which is involved in tissue-specific gene regulation [60] and expressed constantly during the keratinocyte differentiation ( S14C Fig ) . The emerin binding protein LMO7 is a cell type-specific transcription factor that is strongly up-regulated during differentiation . It acts by escaping actin ( ACTB ) -mediated inhibition [61 , 62] and therefore its levels need to increase strongly . At the same time , Laminin B1 ( LMNB1; important for DNA replication ) is transiently down-regulated . Finally , the BAF ( SWI/SNF ) complexes are known for their combinatorial assembly providing functional specificities [63 , 64] ( S14D Fig ) . BRCA1 can directly interact with the BRG1 subunit of the SWI/SNF complex and is down-regulated during differentiation . This may liberate the SWI/SNF complexes to take part in chromatin remodeling , which are either constantly expressed or moderately down-regulated . Generalizing , there are three different types of assembly motifs: In some complexes , dynamic genes are added or removed from the complexes and these are predicted to be compatible ( AND; [14] ) interactions ( e . g . TNF-α/NFκB signaling and Notch complexes ) . For other complexes we observed opposing expression changes of subunits , which are potentially competing for the same binding interface ( XOR interactions; i . e . CULIN SKP2 CAND1 [14] ) . Finally , there are large assemblies where we observed a mixed behavior ( i . e . EGFR/TGF-β; S16 Fig ) . In summary , we find a di-chromatic behavior with a stable core for many complexes involved in cell differentiation . We suggest that the different assembly motifs with respect to compatible ( AND ) and mutually exclusive ( XOR ) surface interactions should be classified using 3D structural information to analyze which type of assembly motifs dominate during keratinocyte differentiation . Proteins that bind mutually exclusively to the same domain on a shared binding partner protein prevent each other’s binding through steric hindrance depending on concentration and localization [10 , 13 , 14] . Steric hindrance and competition could be a mechanism to achieve cell type-specific functions if a competing protein is expressed at a higher level in a specific cell type or tissue [14] . To determine if the replacement of subunits of complexes during the differentiation process happens at mutually exclusive surface interactions , we structurally analyzed all CORUM complexes with less than 20 members using the SAPIN software framework ( S17 Fig ) [19] . SAPIN identifies the protein regions that could be involved in an interaction , provides template structures , and then performs structural superimpositions to identify compatible and mutually exclusive interactions . If a protein has at least two interacting partners , the domains mediating the interaction are superimposed on the reference domain , and the interacting domains are analyzed for compatibility ( AND ) or mutual exclusiveness ( XOR ) ( S4 Table ) . Next , we combined expression classification during keratinocyte differentiation ( non-dynamic vs . dynamic ) with compatible and mutually exclusive interaction types ( S18 Fig ) . Out of six possible cases ( obviating the unresolved group ) , three cases were selected that we could interpret in terms of competition ( Fig 6A ) . In case 1 , the hub protein ( common protein with at least two different binding partners ) is non-dynamically expressed while the attachment proteins are dynamic . In case 2 , all three interacting proteins are dynamic . In case 3 all three interacting proteins are non-dynamic . Interestingly , case 1 is significantly enriched for gene products with mutually exclusive surface interactions ( XOR ) ( 61% compared to 39% , Fisher’s exact test; P = 1 . 9e-6 ) , reinforcing our hypothesis that dynamic genes tend to be involved in competing interactions for a common ( constitutively non-dynamically expressed ) binding partner ( Fig 6 and S19–S21 Figs ) . Case 3 is significantly enriched for AND ( 61% compared to 39% , Fisher’s exact test; P = 2 . 03e-12 ) , representing stable complexes that do not change their composition during the differentiation process . Case 2 ( all three proteins dynamically changing ) is also enriched for XOR interactions compared to AND ( 65% compared to 35% , Fisher’s exact test; P = 8 . 66e-04 ) . When focusing on the 32 XOR nodes with at least two super-dynamic genes , 26 XOR nodes contain genes from at least 2 different clusters suggesting that opposing or at least different expression profiles may impose different functional outputs possibly to compete at single interface hubs . Thus , variation in the concentration of proteins ( belonging to different clusters ) that bind to XOR nodes may cause complex re-assembly through competition and thus achieve a different functional output in the differentiated keratinocytes or during the process of differentiation .
Using calcium-induced differentiation of primary keratinocytes as a model system for a substantial cellular re-organization , we analyzed transcriptome changes during 45 hours . We discovered a large proportion of genes change their expression during the differentiation process ( 36% dynamic and 7 . 9% super-dynamic genes ) , which is comparable with other cell differentiation or reprogramming studies [2 , 4 , 65] . Our set of dynamically up-regulated genes contains 39 out of 53 previously described keratinocyte differentiation markers , 25 of which are super-dynamic [66] . The super-dynamically changing genes , aside from those with known keratinocyte function , span metabolism [67] , signaling , and housekeeping cellular functions . Interestingly a quarter of super-dynamic genes are of unknown functions based on UniProt annotations and manual literature searches . The super-dynamic genes were partitioned into eight clusters , with only those genes needed for establishing the final physiological functions of the cornified envelope ( e . g . water repellence , structural stability , and mechanical resistance ) exhibiting a clear GO-enrichment . Many metabolic , signaling and housekeeping genes were found in all clusters . Interestingly , the shapes of the clusters are similar to those observed in other cell differentiation or reprogramming studies [4 , 65] . Yet , due to an absence of strong enrichment in GO terms in this study and others , direct comparison between the functions of the clusters could not be conducted . Both correlated and anti-correlated paralogous pairs have been discovered before , during reprogramming of somatic cells to pluripotency [4] . Similarly , we found that dynamic paralogs are enriched in correlated expression changes , which may represent examples of gene duplicates being maintained to increase gene dosage/expression levels [31–33] . In addition , gene duplication has also been shown to contribute to the robustness of complex formation [68] . We also found that dynamically changing paralogs can be anti-correlated . Here we showed that anti-correlated paralogous pairs have a greater amino acid sequence and Pfam domain divergence and are evolutionary older genes , suggesting that paralogs have attained specific roles during differentiation . To analyze how dynamic and non-dynamic genes are integrated into complexes , we used the CORUM database to map the changes in gene expression onto protein complexes . Earlier work analyzing complex assemblies during the yeast cell cycle revealed that complexes consist of both constitutively non-dynamic and dynamically expressed subunits ( ‘just-in-time assembly’ ) [6] . Our study of a mammalian differentiation process augments these efforts as we find a larger fraction of di-chromatic complexes , containing both dynamic and non-dynamic genes . When we applied these concepts to a manually annotated large signaling complex to include directionality , the TNF-α/NF-κB signaling , we found a stable core associated to both a dynamically changing module and several stable modules . We propose this as a ‘constant signalosome ready to work’ , where a stable core—involving TNF receptor/TRAF2—associates with a dynamic periphery with different functions: i . e . NF-κB signaling , TRAF1 resulting in anti-apoptotic effects without affecting TRAF2/TRAD . On the level of EGFR signaling , a picture emerges whereby several sub-complexes consisting of adaptors , Raf kinases , ERK kinases , etc . show a mix of both , dynamic and non-dynamic subunits . In contrast to what is proposed in the hour-glass model of signaling [69 , 70] , our analysis suggests that several sub-complexes act in parallel , containing both common non-dynamic and different dynamic subunits . An alternative view of complexes ( ‘machines vs . ensembles’ ) has been proposed by Loew and colleagues , which holds that in addition to stable signaling complexes and molecular machines , such as the ribosome or the proteasome , a considerable combinatorial complexity arises from different compositions of complexes [8 , 71 , 72] . We propose that these two concepts do not exclude each other and are probably both used in signal transduction . However , it is currently unclear how ‘ensemble-like’ signaling complexes actually are . Both , computational and experimental work is critical to answer this question and in particular structural analyses help to identify the two cases by distinguishing compatible ( more machine-like ) from mutually exclusive ( more ensemble-like ) complexes [10 , 71] . Structural information about protein domains was used to distinguish compatible and mutually exclusive protein-protein interactions among hub proteins and their binding partners [10 , 19] . Here , we used the SAPIN webserver to distinguish compatible from mutually exclusive binding surfaces [19] . Other successful methods for structural characterization are based on evolutionary conservation of the interacting residues , e . g . PrePPI [73] , Inferred Biomolecular Interaction Server—IBIS webserver [74] , and Interactome3D [15] . We identified that dynamic genes binding to a common non-dynamic hub are enriched in mutually exclusive interactions ( XOR ) , suggesting that changes in complex assemblies during differentiation could also be caused through variation in the abundance of proteins that compete for binding to XOR nodes [14] . In fact we show here that XOR nodes involving super-dynamic genes are often from different clusters , supporting this hypothesis . Thus we go one step further and propose a model where at least some protein complexes exist assembled stably at the core , and their behavior is modified by the attachment and detachment of accessory proteins at the periphery of the signaling cascade in response to various conditions . This is in agreement with the concept of achieving cell type-specific complexes through the interaction of core ( cell/tissue general ) and peripheral ( cell type/tissue-specific ) proteins [7 , 75] . However , more detailed computational modeling , including information on XOR and AND nodes are needed to investigate the global impact on protein competition , e . g . how changes in protein abundance propagate through a larger PPI network [71] . We have found examples where anti-correlated changes in gene expression of subunits in a complex ( Ubiqutin/degradation ) , or in two complexes competing for the same component ( EGFR recycling complexes ) , which reflect very nicely what is known about factors important for keratinocyte differentiation . This supports the idea of protein competition as a driver of biological processes . In fact , computational modeling of the yeast PPI network has suggested that changes in concentration spread locally and decrease exponentially within the network , as a function of the distance from the perturbed node [76] . Previous experimental and computational work investigating competition at the RAS node has confirmed these rather local effects of competition [14] . Furthermore , it is known that competing protein interactions can induce switch-like cellular behaviors , such as apoptosis versus differentiation [77] , or self-renewal versus differentiation [78] . However , cellular fates also crucially depend on the crosstalk between signaling pathways initiated at different phosphorylation sites providing spatiotemporal separation and acting as molecular switches [79] . As our gene expression changes did not allow us to analyze phosphorylation events , it is difficult to speculate to what extend phosphorylation levels contribute to network and complex reorganization processes during keratinocyte differentiation . Likewise , the role of homooligomeric complexes is neglected despite being the dominating type of interaction . In the future , it will be important to integrate protein interaction networks resolved at the level of domains as well as phosphorylation events and homooligomeric complexes , in order to provide the complete picture . Altogether our analysis highlights the importance of understanding the dynamic assembly and disassembly of complexes taking 3D structural information into consideration , rather than unraveling networks of individual proteins .
The time-course microarray data analyzed in this work was generated in our previous study [3] . However , in this section of the methods , we re-outline the details of this experiment . Total RNA in the amount of 500 ng from calcium treated primary human keratinocytes collected every 5 hours in triplicates for a total of 45 hours ( i . e . 10 samples ) and labelled using Agilent LowInputQuick Amp Labelling kit following manufacturer instructions . mRNA was reverse transcribed in the presence of T7-oligo-dT primer to produce cDNA , The cDNA was then in vitro transcribed with T7 RNA polymerase in the presence of Cy3-CTP to produce labeled cRNA and this labeled cRNA was in turn hybridized to Agilent Human Gene Expression 4x44K v2 microarrays ( ID026652 ) . Signals from probes were obtained by the Agilent’s Feature Extraction custom software , were corrected for background noise using the normexp method [80] available in the R package limma from Bioconductor , and then normalized between arrays to assure comparability across samples using quantile normalization [81] . Lastly , the dataset was log2-transformed . The dataset was filtered to remove and collapse all Agilent control spots and to only include probes that were present above stochastic background expression . To do this we filtered any probes with expression profiles completely below 7 which is a figure obtained by taking the median of the entire dataset and rounding it down to a single digit . This procedure yielded 21 , 113 probes , which were then mapped to 16 , 720 genes where the probe with the highest mean expression represents a gene . For some analyses such as K-means clustering the dataset was mean-centered and scaled such that the mean of each gene is zero . To distinguish genes that are changing ( i . e . transiently or continuously up- or down-regulated ) from those that are constant ( i . e . stably expressed ) across our differentiation time-course , we used Chi-squared test where for each gene , the expected value at each time point is equivalent to the gene’s average expression across all 10 . The equation is as follow: X2=∑iN ( X¯i−X¯ ) 2N−1 where N is the number of samples or time points ( i . e . 10 ) , X¯i is the mean expression of gene x across three experimental repeats at time point i , and X¯ is the overall mean expression of gene x across all 10 time points and all repeats . Note that in this formula there is no normalization of ( X¯i−X¯ ) 2 by the experimental error computed as SE = SD/N . Instead we computed a universal error and scaled the X2 distributions of our data to the standard X2 distribution for 9 degrees of freedom . In addition to Chi-squared test we employed an empirically derived fold change threshold of 2 and a strict threshold of 4 . Each gene’s expression peak was compared to its trough to ensure the difference satisfies the appropriate threshold and the genes were classified as follow: non-dynamic genes have to satisfy P ( X2 ) > = 0 . 01 and fold change < 2 unresolved genes either have P ( X2 ) <0 . 01 and fold change < = 2 or P ( X2 ) > = 0 . 01 and fold change > = 2 dynamic genes P ( X2 ) <0 . 01 and fold change> = 2 . super-dynamic genes are a subset of dynamic genes which have fold change> = 4 We also conducted corrections for multiple testing using Benjamini’s method . While the above analysis had classified 6 , 096 genes as dynamic ( 1 , 317 as super-dynamic ) , after Benjamini correction for multiple testing and applying a p-value cutoff of 0 . 01 ( 1% ) , these two numbers remained the same while the number of non-dynamic genes increased from 6 , 137 to 6 , 413 and the number of unresolved genes decreased from 4 , 487 to 4 , 211 . Thus , a subset of 200 genes got shifted from the unresolved category to the non-dynamic category after Benjamini correction . Since the majority of our analyses were based on dynamic and super-dynamic genes such as clustering and mapping to complexes and these classes did not change after multiple testing correction , we kept the gene classes as originally defined without Benjamini correction . We used R version 2 . 14 for the majority of the statistical analyses along with Perl and AWK for text processing . K-means clustering was performed using the R kmeans function . To optimize and select the K value or the number of cluster centers , we exhaustively performed K-means clustering with K value ranging from 2 to 500 and then for each set of clusters , computed an F-score as well as an average silhouette score available from the standard statistics and cluster packages , respectively . The K value , which resulted in the highest average silhouette score and F-score , was selected as the optimal number of cluster centers . A reference set of complexes was obtained from the CORUM database [36] of curated mammalian protein complexes . Out of a total of 1 , 331 human complexes available in the latest release ( February 17th , 2012 ) , 44 homo-oligomer complexes consisting of only one type of protein were filtered out . 84 complexes were removed , which were fully non-expressed across our skin differentiation expression time course . 154 ( <13% ) complexes contain subunits that are partially non-expressed in keratinocytes while 1 , 049 complexes contain subunits , which are all expressed during skin differentiation . The resulting set of 1 , 203 complexes consists of 2 or more distinct proteins per complex . Paralog annotations were obtained from EnsemblCompara through the BioMart portal ( database version EnsemblGenes71 ) . The method identifies true paralogs by computing a phylogenetic tree across the whole set of protein-coding genes with one pipeline which includes TreeBeST [33] . Analysis of mutually exclusive ( XOR ) and compatible ( AND ) binding was done using the SAPIN web framework [19] . SAPIN identifies the protein parts that could be involved in the interaction and provides template structures and then performs structural superimpositions to identify compatible and mutually exclusive interactions . We analyzed all complexes in the CORUM database with a complex size of less than 20 members . Each CORUM complex was broken down into combinations of three different proteins , and the respective protein sequences were used as the input for the SAPIN webserver . The workflow in SAPIN is: Assigning all domains using Pfam ( pfam_scan . pl with default parameter ) based on the protein sequences . Searching the 3DID [82] database to find potential domain-domain interaction hits of binary interaction partners provided . Finding the best template of PPI structures ( containing interacting domain A and domain B ) by using the InterPreTS scoring function [83] . The interaction structures are evaluated by InterPreTS based on interface sequences aligned with 3Dstructures ( we used MUSCLE [84] with default parameter ) . SAPIN selects one template that has more than 2 . 33 Z-score or the best score one among all candidates . If two PPI structures share a same domain , structural superimpositions are performed based on the domain structure using Combinatorial Extension ( CE [85] ) . Analyzing for backbone clashes using FoldX [86] with superposed structures , and based on a backbone clashes threshold ( 15% of interface residues ) the interactions are assigned compatible ( AND ) or mutually exclusive ( XOR ) . The reliability of the predictions with respect to sequence similarity to the template complex , we measured the sequence identity between the reference and homolog domains from the alignment based on the Needleman-Wunsch Algorithm [87] . Based on this , we determined a Z-score for the percentage of van der Waals backbone clashing among interface residues calculated for either AND or XOR , dependent on sequence similarity [19] . If a protein has at least two interacting partners , the domains mediating the interaction are superimposed on the reference domain , and the interacting domains are analyzed for compatibility ( AND ) or mutual exclusiveness ( XOR ) . Note , that even two protein having similar domains do not necessarily bind at the same site/interface , but there are frequent cases of similar domains binding in a different way . As the best template suggestion based on sequence homology using INTERPRETS [9] . As the sequence is scored in INTERPRETS against a set of 1000 random sequences , the selection of the best template is not deterministic and can result in different results in different runs . We provide the average of the XOR/AND likelihood of independent runs in S4 Table . For calcium-induced differentiation experiments , keratinocytes were seeded into 35 mm plates and grown in Keratinocyte Serum-Free Medium with supplements ( KSFM; GIBCO ) [3] . After reaching 70% confluence Keratinocyte Serum-Free Medium was exchanged for EMEM ( Lonza ) supplemented with 8% chelated FBS , EGF ( 10 ng/ml ) , 1% penicillin/streptomycin and 0 . 05 mM CaCl2 . After 12 hour time point 0 h was collected , and the residual keratinocytes were synchronized for 2 h with EMEM containing 20% chelated FBS , EGF ( 10 ng/ml ) , 1% penicillin/streptomycin , and 0 . 05 mM CaCl2 . After synchronization cells were washed once with PBS and cultured in EMEM , supplemented with 8% chelated FBS , EGF ( 10 ng/ml ) , 1% penicillin/streptomycin , and either 0 . 05 mM or 1 . 2 mM CaCl2 , corresponding to non-calcium and calcium treatment , respectively . Cells were collected every 5 hours during a period of 45 hr . Keratinocytes were seeded onto 16-chambered LabTek slides ( Nuncbrand ) . Kerationocyte differentiation was induced through addition of Calcium ( as before ) , and sample were taken at the respective time points . Cells were fixed in 4% paraformaldehyde for 20 min and blocked with 4% BSA in PBS 1× ( blocking solution ) . Primary staining was done using an antibody against Involucrin ( Abcam , ab28057; dilution 1:1000 in blocking solution ) followed by an Alexa Fluor 568 secondary antibody ( Invitrogen Probes; dilution 1:200 in blocking solution ) . Cells were counterstained with DAPI ( Sigma-Aldrich; concentration 1 μg/mL ) and mounted using Mowiol solution . Staining was visualized on a Leica TCS SP5 confocal microscope with a 40× 1 . 25 NA objective at a zoom factor of 3 ( 1024 × 1024 pixels; 0 . 126 μm/pixel ) . | A key challenge in cellular network biology is to understand how protein complexes are cell-type or condition-specific assembled and disassembled . Cell differentiation is a major cellular reorganization bringing about fundamental changes in the new differentiated cell type . As many genes are expressed throughout all stages and only their expression levels differ , the question arises of how specific functions can be mediated . Here , focusing on the calcium-induced differentiation of primary human keratinocytes , we describe motifs of protein complex assemblies . We found that a large proportion of complexes contain both proteins expressed at similar levels in all stages of differentiation ( non-dynamically expressed ) and proteins with variable expression between ( dynamically expressed ) . Using structural information we found that subunits tend to be replaced at structural overlapping surfaces of proteins . When applying our concepts to a manually annotated large TNF/NFkB signaling complex , we find a stable core associated with both a dynamically changing module and several stable modules . We propose this as a ‘constant signalosome ready to work , ’ where a stable core is associated with a dynamic periphery . Altogether , our analysis highlights the importance of understanding the dynamic assembly and disassembly of complexes , taking 3D structural information into consideration , rather than only considering networks of individual proteins . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Dissecting the Calcium-Induced Differentiation of Human Primary Keratinocytes Stem Cells by Integrative and Structural Network Analyses |
The genetically tractable model host Caenorhabditis elegans provides a valuable tool to dissect host-microbe interactions in vivo . Pseudomonas aeruginosa and Staphylococcus aureus utilize virulence factors involved in human disease to infect and kill C . elegans . Despite much progress , virtually nothing is known regarding the cytopathology of infection and the proximate causes of nematode death . Using light and electron microscopy , we found that P . aeruginosa infection entails intestinal distention , accumulation of an unidentified extracellular matrix and P . aeruginosa-synthesized outer membrane vesicles in the gut lumen and on the apical surface of intestinal cells , the appearance of abnormal autophagosomes inside intestinal cells , and P . aeruginosa intracellular invasion of C . elegans . Importantly , heat-killed P . aeruginosa fails to elicit a significant host response , suggesting that the C . elegans response to P . aeruginosa is activated either by heat-labile signals or pathogen-induced damage . In contrast , S . aureus infection causes enterocyte effacement , intestinal epithelium destruction , and complete degradation of internal organs . S . aureus activates a strong transcriptional response in C . elegans intestinal epithelial cells , which aids host survival during infection and shares elements with human innate responses . The C . elegans genes induced in response to S . aureus are mostly distinct from those induced by P . aeruginosa . In contrast to P . aeruginosa , heat-killed S . aureus activates a similar response as live S . aureus , which appears to be independent of the single C . elegans Toll-Like Receptor ( TLR ) protein . These data suggest that the host response to S . aureus is possibly mediated by pathogen-associated molecular patterns ( PAMPs ) . Because our data suggest that neither the P . aeruginosa nor the S . aureus–triggered response requires canonical TLR signaling , they imply the existence of unidentified mechanisms for pathogen detection in C . elegans , with potentially conserved roles also in mammals .
The study of host-microbe interactions seeks to understand the symbiotic relationships between hosts and microbiota , and their perversion during infectious disease . Essential steps are the identification of bacterial virulence mechanisms and of host defense pathways . In mammalian hosts , Nod-like receptors ( NLRs ) , Toll-like receptors ( TLRs ) , and NF-κB play important roles in the intestinal epithelium , a critical interface of contact between host and microbiota [1] , [2] , [3] . However , how these signaling pathways function in the context of the whole organism is poorly understood , and potentially novel pathways may yet be uncovered . Likewise , the critical initial stages of infection , before the onset of overt pathogenesis , are poorly defined . Genetically tractable invertebrate model systems have aided efforts to identify evolutionarily conserved components of the innate immune system [4] . For example , studies using Drosophila melanogaster showed the central importance of the Toll and IMD signaling pathways for the regulation of Relish-family ( NF-κB ) transcription factors [3] , [5] . Likewise , studies using Caenorhabditis elegans revealed the involvement of evolutionarily conserved p38 MAPK , insulin , TGF-β , and β-catenin signaling pathways [6] , [7] , [8] . In addition to being a genetically tractable model system , C . elegans is a transparent bacterivore , which allows the direct , real-time observation of infection and gene expression reporters in vivo . These qualities make it a useful model host for the study of infection and host defense in the context of the whole organism [9] . C . elegans is particularly useful for studying intestinal epithelial innate defenses , because it has only 20 such cells that are not shed ( as are mammalian intestinal epithelia ) and are non-renewable [9] , [10] , allowing the study of defense functions in vivo without potentially confounding cell proliferation and tissue repair . Furthermore , the unique biology of C . elegans allows researchers to focus entirely on epithelial innate defense because it lacks a circulatory system , macrophage-like professional immune phagocytes , and antibody-based adaptive immunity [11] . On the bacterial side , it is important to elucidate the virulence mechanisms that defeat host defenses and establish infection . Pathogenic bacteria are thought to have experienced stepwise additions of virulence factors , as they evolved to survive different host antimicrobial responses and to colonize new niches [12] . Our studies using C . elegans as a model host may thus interrogate early steps in the evolution of bacteria as pathogens , and their interactions with prototypical metazoan epithelial cells . Here we focus on two paradigmatic human pathogenic bacteria of great medical importance that represent two broad categories of evolutionarily distant microbes , the Gram-negative Pseudomonas aeruginosa and Gram-positive Staphylococcus aureus . P . aeruginosa causes systemic acute infections in patients with weakened immune systems [13] and establishes chronic infections in the lungs of cystic fibrosis patients [14] . P . aeruginosa can also infect a wide variety of plants , metazoans , and single-celled eukaryotes [15] . S . aureus is a Gram-positive bacterium that can cause severe diseases in many animal species [16] , [17] . In recent years , patients lacking classical risk factors have suffered increasing rates of infection by virulent antibiotic-resistant strains [18] . Human colonization by S . aureus is widespread: 30% of the population carries S . aureus in the microflora of epithelia in the nasopharynx , skin , and intestine [19] . S . aureus can cause severe skin infections , osteomyelitis , endocarditis , food poisoning , pneumonia , and flesh-eating disease [20] , for which it deploys an impressive armamentarium of virulence factors , including cytolysins that cause the destruction of host immune cells and tissues [21] , [22] . Despite great progress in their identification , the exact contribution of each virulence strategy to disease in vivo is poorly understood . The genetic makeup of the host is suspected to determine susceptibility to infection , but the genetic determinants of susceptibility are unknown [20] . Mice lacking adaptive immunity survive intravenous S . aureus infection as well as wild-type animals , suggesting that innate immunity is the main clearing mechanism for S . aureus infection in mammals , but the exact mechanism is unclear [23] . New approaches are needed to understand the molecular basis of innate host defenses against P . aeruginosa and S . aureus infection . To this end , our laboratory has developed C . elegans-P . aeruginosa [24] , [25] and C . elegans-S . aureus [17] , [26] model systems to facilitate the study of the role of innate host defenses in conferring resistance to bacterial infections and to identify host signaling pathways relevant to defense [7] , [27] , [28] , [29] . These infection models recapitulate key aspects of P . aeruginosa or S . aureus disease in mammals ( see below ) , including the requirement of virulence factors necessary for mammalian infection , and have been used to identify novel P . aeruginosa and S . aureus virulence factors [17] , [30] , [31] , [32] . Despite great progress in the dissection of C . elegans host defense signaling pathways since the initial description of the system in 1999 [9] , [24] , [33] , [34] , little information has been available on the cellular basis of bacterial pathogenesis and nematode killing . In this study , we focused on the interactions between C . elegans intestinal cells —as prototypical metazoan epithelial cells , and as the first line of defense against intestinal infection—and P . aeruginosa or S . aureus . We investigated the cytopathologies that occur during infection , which suggest distinct mechanisms of virulence used by each bacterial species in vivo . With P . aeruginosa , we found that initial intestinal distention , putative outer membrane vesicle ( OMV ) production , and extracellular matrix accumulation on the intestinal cell brush border are followed by host autophagic abnormalities , intracellular invasion , and penetration of the epithelial barrier . Similarly , previous studies found that P . aeruginosa forms biofilms in the lungs of infected patients , where OMV production is also evident [35] , [36] . In contrast , faster accumulation of S . aureus in the C . elegans intestine resulted in enterocyte effacement and loss of intestinal cell volume , followed by intestinal epithelial cell lysis and bacterial invasion of the rest of the body , with complete degradation of host internal tissues . Likewise , previous studies showed intestinal colonization by S . aureus , enterocyte effacement in rabbits and neonates , and toxin-mediated cell lysis both in vitro and in vivo [37] , [38] , [39] , [40] , [41] , [42] , [43] . We also evaluated the differential impact of these distinct pathogenic processes on host gene transcription . We previously defined the host transcriptional response to P . aeruginosa infection [44] . To understand if and how the host responds to different virulence mechanisms by employing distinct transcriptional host responses , here we defined the host response to S . aureus . The two responses show minimal overlap; the response to S . aureus apparently involves host damage- and TLR-independent recognition of microbial molecules , potentially pathogen-associated molecular patterns ( PAMPs ) , whereas C . elegans may sense P . aeruginosa-derived heat-labile signals or pathogen-elicited damage . Using functional genomics , we identified host factors critical for host defense against S . aureus , some of which are analogous to human innate defense factors . These observations advance our knowledge of bacterial pathogenesis in C . elegans , and show that the C . elegans infection model illuminates evolutionarily conserved mechanisms of bacterial pathogenesis and epithelial host defense .
To determine the cytopathology of P . aeruginosa colonization of the C . elegans intestine , we used transmission electron microscopy ( TEM ) of the intestinal epithelium ( Figure 1A ) to evaluate signs of pathogenesis at early times ( 8 h ) or at later times ( 24 h and 48 h ) of infection . At 8 h of infection , we found gross intestinal distention but little bacterial accumulation . Instead , we observed unidentified electron-dense extracellular material accumulating on the apical surface of the brush border ( Figure 1C ) . In addition to coating the brush border , the electron-dense material surrounded bacterial cells that appeared intact and formed clumps in the intestinal lumen . Also surrounding the bacteria and in contact with the extracellular material , we found abundant accumulation of putative outer membrane vesicles ( OMVs ) ( Figure S1A , B ) . P . aeruginosa OMVs have been shown to act as a virulence factor and toxin delivery mechanism [45] . We did not observe intestinal distention , OMVs , or matrix accumulation in E . coli-fed control animals ( Figure 1B ) . At 24 h of infection , the intestine became further distended , with noticeably more bacterial cells accumulating in the lumen in the form of clumps of cells surrounded by extracellular matrix ( Figure 1E ) . There was a thick layer of matrix material coating the microvilli , which were present and of approximately normal length ( i . e . , ∼1 µm ) . In contrast , E . coli-fed animals lacked these signs of pathogenesis , exhibiting non-distended intestinal lumina and intestinal epithelial cells filled with lipid droplets and other gut granules characteristic of healthy animals ( Figure 1D ) . At 48 h of infection , pathogenesis advanced further , resulting in higher levels of bacterial accumulation in the grossly distended intestinal lumen ( Figure 1G ) . The bacterial cells were mostly not in direct contact with the microvillar surface , but separated from it by a thick layer of extracellular material . At this time , there was widespread shortening of the microvilli and intracellular invasion by the bacteria ( Figure 1G ) . Intracellular invasion was observed in 21% of cross sections ( N = 14 ) , only after 48 h infection . In some cases , we found bacterial cells at distal sites beyond the intestine , suggesting that P . aeruginosa can penetrate the intestinal cells and invade other tissues ( Figure S1C ) . In addition to these phenotypes , we observed an increased number of autophagosomes , readily identifiable by their multi-membranous structure ( Figure 1H and S2 ) . Indeed , most autophagosomes appeared to be either early autophagosomes ( Figure 1H ) or aberrant multivesicular autophagosomes ( Figure 1I ) . In contrast , mutant gacA P . aeruginosa , lacking the master virulence regulator GacA and therefore attenuated for C . elegans killing [25] , caused much lower levels of autophagosome accumulation ( Figure S2 ) , pathogenesis ( Figure 1F ) , and OMVs ( Figure S1D ) by 48 h . Intracellular PA14 gacA was not observed , even after 72 h ( N = 15 ) . We observed less dense matrix accumulation in gacA mutant-infected animals than with wild type P . aeruginosa , and did not observe microvillar shortening , intracellular invasion , or severe luminal distention . At all times during the infection by both wild-type and gacA P . aeruginosa , we only observed what appeared to be live P . aeruginosa cells , in contrast to S . aureus as described below . Interestingly , the cytopathology of S . aureus infection in the C . elegans intestine is markedly different from a P . aeruginosa infection . First , using GFP-labeled S . aureus , we observed rapid accumulation of bacteria in the intestine 4 h after infection initiation , whereas P . aeruginosa did not start accumulating until 8 after initial exposure ( Figure S4A , B ) . At 4 h , S . aureus accumulated in the anterior and posterior ends of the intestine , and the rectum ( Figure S3A , C ) , with less accumulation in the mid section of the intestinal lumen , where the bacteria appeared to be adhering to the apical surface of intestinal cells ( Figure S3B ) . S . aureus accumulated further over the course of the following 4 h ( Figure S4B ) . The infected animals moved slowly , were smaller ( Figure S4C , D ) , and appeared to produce fewer eggs , than healthy animals . In addition to the intestinal distention and accumulation phenotypes , we observed a marked deformation of the anal region with S . aureus ( Figure S4C ) but not with P . aeruginosa ( not shown ) or non-pathogenic E . coli ( Figure S4D ) . This deformed anal region ( Dar ) phenotype [46] appeared 4–8 h after initiation of infection and required live S . aureus ( Figure S4E ) . Interestingly , this Dar phenotype was dependent on bar-1/β-catenin and mpk-1/extracellular signal regulated kinase ( ERK ) ( Figures S4F , G , J , K , and L ) , which is also required for the Dar response to Microbacterium nematophilum [47] . We previously showed that bar-1/β-catenin and its downstream target gene egl-5/HOX exhibit a defective intestinal response to S . aureus [7] . Unexpectedly , mutants defective in egl-5/HOX exhibited a wild-type Dar response ( Figure S4H , L ) , despite having an altered intestinal host response to S . aureus [7] and a defective anal swelling response to M . nematophilum infection [48] . Similarly , pmk-1/p38 MAPK mutants exhibited a slightly less noticeable , but equally frequent , Dar phenotype following S . aureus infection ( Figure S4I , L ) , consistent with our previous observation that pmk-1 mutants are only subtly more susceptible to S . aureus-mediated killing [7] . These data suggest that the Dar phenotype may be a defensive host swelling response to pathogen-mediated host damage , since it requires an active host response to live bacteria . To investigate the cytopathology of S . aureus infection , we performed TEM of S . aureus-infected animals , focusing on the intestinal epithelium ( Figure 2A ) . After 12 h of infection , we found a striking decrease in the length of the microvilli compared to animals feeding on non-pathogenic E . coli ( Figures 2B–E , 3B–C ) . We also observed significant plasma membrane “blebbing” from the apical surface of intestinal cells ( Figures 2D–E , 3C , S6A–B ) . The intestinal lumina of S . aureus-infected animals were markedly distended , consistent with our previous observations using light microscopy [17] . Distention was apparently a consequence of severe volume loss of the intestinal epithelial cells , with concomitant accumulation of bacterial cells in the enlarged luminal space ( Figure 2B , D ) . In marked contrast to P . aeruginosa , an average 34% of S . aureus cells in the lumen ( 9–63% , N = 8 cross sections ) were lysed at 12 h of infection ( Figure 2E ) ; these appeared similar to published TEM micrographs showing S . aureus cells killed with antimicrobial peptides in vitro [49] , suggesting that the C . elegans intestine may produce bactericidal factors active against S . aureus . In contrast to animals feeding on E . coli , at 24 h of infection the microvilli were almost completely destroyed ( Figure 3D , E ) and at 36 h were completely absent , in what is termed “enterocyte effacement” ( Figure 3F , G , I ) . Further , at 36 h we observed a reduction of intestinal cell volume ( see thin sliver in Figure 3G ) and intestinal cell lysis ( Figure 3I ) . Also at 36 h , we observed a few dead animals , the organs of which were completely degraded except for the collagenous cuticle and an unidentified circular internal structure ( Figure S5 ) . Heat-killed S . aureus did not cause intestinal distention , microvillar effacement , intestinal cell lysis , or death ( Figure 3H ) . These data show that S . aureus causes membrane and cytoskeletal rearrangements , as well as enterocyte effacement and destruction , possibly by secreted membrane-active bacterial toxins such as cytolysins or other pore-forming toxins ( PFTs ) . Hemolysins α , β , and γ are known S . aureus cytolysins . However , they appeared to not be required for pathogenesis and killing , as a S . aureus strain lacking all three hemolysins exhibited similar kinetics of C . elegans killing as the isogenic wild type ( Figure S7 ) . Similarly , the α-hemolysin Δhla mutant was as capable of causing enterocyte effacement , intestinal distention , membrane blebbing , and intestinal cell lysis as wild type ( Figure S6C–F ) . These results indicate that virulence factors other than the hemolysins are responsible for the observed intestinal cell lysis . Because host defense from , and digestion of , ingested bacteria are necessarily linked in bacterivorous animals such as C . elegans , the distinction between innate immune responses and digestive responses is blurred . Previous studies have investigated the long-term effects of ingestion of pathogenic bacteria , defining a common necrotic host response that is triggered by several pathogens after 24 h of infection [50] . To investigate gene expression changes more likely to be elicited directly by S . aureus detection , we evaluated gene expression at an earlier infection time , namely 8 h . We previously reported that P . aeruginosa induces a potent transcriptional host response early during infection of C . elegans , which significantly contributed to our understanding of C . elegans defense from P . aeruginosa infection [44] . To determine whether C . elegans mounts a similar host response to S . aureus infection , we performed whole-genome transcriptional profiling of animals infected with S . aureus for 8 h , relative to animals feeding on non-pathogenic E . coli . We found 186 transcripts that increased at least two-fold in abundance and 198 that decreased at least two-fold after infection ( Table S1 ) . Focusing on 46 genes up-regulated 4-fold or higher as a smaller sample , we found that the majority had potential xenobiotic detoxification or antimicrobial activities , consistent with their involvement in a protective host response ( Table 1 ) . In this group , a number of genes of unknown function appeared to encode short secreted polypeptides that may possess antimicrobial activities ( Table 1 ) . To identify potential physiological roles of this host response , we used two complementary methods to study the over-representation of gene ontology ( GO ) classes ( see Materials and Methods ) . These analyses revealed up-regulation of detoxifying and antimicrobial responses and down-regulation of growth-related metabolic pathways and extracellular structural components ( Table S7 ) . Significance of representation analysis revealed that the most significantly induced gene class contains sugar-binding proteins including C-type lectins ( CTLs , N = 15 , p = 1 . 2E-5 ) , which could act as signaling receptors , opsonizing agents , or direct antimicrobial effectors [51] , [52] , [53] , [54] . The most significantly repressed GO classes contain genes encoding structural constituents of the cuticle ( e . g . collagens; N = 47 , p = 4 . 4E-25 ) , phosphate and inorganic anion transport ( N = 47 , p = 4 . 5E-25 and p = 5 . 8E-21 ) , basement membrane components ( N = 4 , p = 1E-6 ) , and lipid transporters ( N = 5 , p = 6E-6 ) . A second approach ( see Text S1 ) expanded these observations to include additional metabolic enzymes and transporters ( Table S2 ) . These analyses highlight the potential role of CTLs as a major immune effector strategy used by C . elegans during infection by S . aureus , and the significant metabolic component of the host response . Because we observed cell membrane rearrangements suggestive of the activity of membrane-active toxins ( Figures 2E , 3C , E , G , I , S6 ) , we hypothesized that part of the host response to S . aureus may also be triggered by PFTs . Indeed , we found that 22 of 422 probe sets up-regulated during exposure of worms to the Bacillus thuringiensis PFT Cry5B [55] were also induced during infection with S . aureus ( Table S3 ) , significantly higher than the 4 probe sets expected by chance alone . These data suggest that the overlapping set of genes shared by the responses to S . aureus and Cry5B may constitute a host response triggered by intestinal cell membrane disruption . Because we observed evidence of intestinal destruction and nutritional deprivation in animals infected with S . aureus , we hypothesized that infected animals might be starving . Previous work identified 18 genes whose expression changed during starvation [56] . In contrast , only one out of 9 previously identified fasting-induced genes ( acs-2 ) and 4 out of 9 fasting-repressed genes ( lbp-8 , acdh-1 , fat-7 , and F08A8 . 2 ) were affected by S . aureus infection . Furthermore , the fasting-induced gene hacd-1 [56] was repressed during infection . These data suggest that the early transcriptional host response to S . aureus infection is minimally impacted by the starvation response . To validate the microarray experiments , we measured transcript levels for ten selected “biomarker” genes by qRT-PCR over a time course of infection . These ten genes , used as models of the larger host response , were chosen to represent different up-regulation levels and functional annotations ( Table 2 ) . All ten genes tested were induced in response to infection ( Figure 4A , B ) . A subset of these biomarkers , fmo-2 ( FMO ) , ilys-3 ( lysozyme ) , cpr-2 ( protease ) , Y65B4BR . 1 ( lipase ) , exc-5 ( GEF ) , and F53A9 . 8 ( putative antimicrobial peptide ) , were already induced 10-fold or higher by the first time-point at 4 h ( Figure 4A ) . A second subset , lys-5 ( lysozyme ) , clec-52 , clec-60 , and clec-71 ( CTLs ) , were only modestly induced by 4 h and exhibited a further increase by 12 h ( Figure 4B ) . Thus , time-resolved gene expression analysis revealed the existence of at least two kinetic groups , defined by their expression levels at 4 h . We also measured transcript levels for 8 genes predicted to be repressed upon infection , confirming reduced expression for 7 of them ( Figure 4C ) . Together , these results confirm the predictive value of the genome-wide profiling . To elucidate the spatial pattern of the host response , we used transgenic animals carrying transcriptional reporters in which the promoters for 5 of the 10 biomarker genes ( clec-52 , clec-60 , F53A9 . 8 , fmo-2 , and ilys-3 ) , as well as clec-70 , an additional CTL gene up-regulated by S . aureus and important for host defense ( see below ) , were fused to GFP . We infected these transgenic animals with S . aureus and compared the intensity and pattern of GFP expression with control animals feeding on E . coli . All the genes tested were expressed at low basal levels in the intestines of the latter ( Figure 5 , left panels ) . After infection with S . aureus , all of the GFP reporters were induced in the intestinal epithelial cells ( Figure 5 , right panels ) . ilys-3 , F53A9 . 8 , clec-52 , clec-60 , and clec-70 were all expressed more strongly in the posterior end of the intestine than the anterior . ilys-3 , fmo-2 , and clec-70 were also induced in the pharynx . Although promoter-GFP fusions like these lack potentially important endogenous regulatory 3′ UTR and intronic sequences , these data are consistent with endogenous RNA localization data from ongoing genome-wide in situ hybridization studies of animals feeding on non-pathogenic E . coli ( NextDB , http://nematode . lab . nig . ac . jp/db2/index . php , and Table 1 ) . Thus , the C . elegans transcriptional host response to S . aureus is primarily localized in the intestinal epithelial cells and , in some cases , additional sites ( Figure S8A–H ) . Up-regulation of the 10 biomarker genes could be a result of cell damage caused by S . aureus , or of microbial detection independent of the inflicted damage . To discriminate between these two scenarios , we measured gene induction during exposure to live S . aureus , which causes pathogenesis , or heat-killed S . aureus , which does not ( Figures 3G–I and S4C , E ) . Unexpectedly , all 10 biomarker genes were induced at least equally well on heat-killed S . aureus as on live S . aureus ( Figure 6A ) . This result suggested that the ten biomarker genes form part of a host response against microbe-derived molecules , possibly pathogen-associated molecular patterns ( PAMPs ) [3] , and not of a damage response . Bacterial cell-wall components , such as LPS and flagellin in Gram-negative bacteria and peptidoglycan in both Gram-negatives and Gram-positives , are common PAMPs in many systems [57] . To test whether Gram-positive cell wall components in general were able to trigger the same S . aureus-induced response , we assayed gene induction in animals feeding on non-pathogenic Gram-positive bacterium Bacillus subtilis compared with E . coli-fed controls . Remarkably , of the ten biomarker genes , only fmo-2 , ilys-3 , and Y65B4BR . 1 were induced by B . subtilis , albeit at much lower levels than with S . aureus ( Figure 6B ) . The remaining 7 genes either were not induced or were repressed by B . subtilis . These results suggest that PAMPs other than shared Gram-positive cell wall molecules may be molecular triggers for the S . aureus-induced host response in C . elegans ( see Discussion ) . How is S . aureus detected ? In fruit flies and mammals , Toll-like receptors ( TLR ) are involved in PAMP detection . C . elegans has a single gene encoding a TLR , tol-1 , which has been shown to be important for avoidance responses to Serratia marcescens and for full induction of antimicrobial peptide abf-2 in response to Salmonella enterica [58] , [59] . However , tol-1 was not required for the induction of any of the 10 S . aureus–induced biomarker genes ( Figure 6C ) . Furthermore , tol-1 mutants were not more susceptible to S . aureus-mediated killing than wild type ( Figure S10 ) . One caveat is that the tol-1 ( nr2033 ) allele used , a deletion that eliminates the cytoplasmic TIR domain necessary for signaling [60] , is viable , thus considered a partial loss of function for viability but a null allele for immune signaling [58] . These results show that tol-1 is most likely not required for the C . elegans host response to S . aureus and suggest that alternative mechanisms may exist for PAMP detection . Since mpk-1/ERK is important for the rectal epithelial cell swelling response to M . nematophilum [47] and for the swelling response to S . aureus ( Figure S4K , L ) , we wondered whether mpk-1 might also be important for the intestinal response to S . aureus . However , mpk-1 animals exhibited no measurable defect in the induction of the biomarker genes relative to E . coli-fed controls ( Figure S9 ) . Therefore , mpk-1 is dispensable for at least part of the intestinal transcriptional response to S . aureus , but not the anal swelling response . In contrast to S . aureus , which either alive or dead triggered the induction of 10 biomarker genes ( Figure 6A ) , heat-killed P . aeruginosa did not trigger the induction of 10 P . aeruginosa-induced biomarker genes ( Figure 6D ) . Together , these data are consistent with the idea that C . elegans may recognize S . aureus infection mainly via TLR-independent PAMP detection , whereas it may recognize P . aeruginosa infection via detection of either Damage-Associated Molecular Patterns ( DAMPs ) , or unidentified heat-labile PAMPs . Host responses to pathogenic attack in plants and animals are remarkably pathogen-specific . In C . elegans , pathogen-specific gene induction has been observed at late times of infection ( i . e . , 24 h ) , when damage and necrosis are apparent [50] . To determine whether pathogen-elicited gene induction at earlier times ( i . e . , 8 h ) is also pathogen-specific , we compared the host response to S . aureus with our previously published study of the early host response to P . aeruginosa [44] . Of 186 genes induced by S . aureus and 259 genes induced by P . aeruginosa , 44 genes were induced by both pathogens , which is more than expected by chance ( Figure 7A , Table 3 ) . We validated the results of the microarray comparison using qRT-PCR; four genes predicted to be specifically induced by P . aeruginosa were either unaffected or repressed by S . aureus ( Figure 7C ) , and four genes predicted to be induced by S . aureus were either unaffected or repressed by P . aeruginosa ( Figure 7D ) . In contrast , four of five predicted overlap genes showed increased expression with both S . aureus and P . aeruginosa ( Figure 7E ) . This suggests that the early host response to pathogen attack involves activation of a “pan-pathogen” response against a broad spectrum of pathogens , as well as a more tailored response that is optimized for the defense against the specific class of pathogens that is causing the infection . To further test the hypothesis of the existence of pathogen-shared and -specific components of the host response , we compared the genes differentially affected by S . aureus infection with previously published profiling of C . elegans infected with M . nematophilum [61] . The comparison of microarray studies independently performed in these separate laboratories may underestimate overlapping gene sets due to the use of different infection , RNA extraction , and data processing methods . Nonetheless , we found a relatively high degree of overlap between the responses to S . aureus and M . nematophilum ( 21 genes out of 186 ) ; 10 genes were induced by S . aureus , M . nematophilum , and P . aeruginosa ( Figure 7A , Table 3 ) . In contrast , 44 ( 68% ) out of 65 genes induced by M . nematophilum were not induced by S . aureus . These data further support the existence of a core , shared response and a specific , tailored response . GO annotations of genes affected by S . aureus or P . aeruginosa also exhibit a degree of specificity . Whereas the S . aureus-induced host response includes many sugar binding proteins , the P . aeruginosa-induced response is not characterized by any particular over-represented GO annotation ( Table S7 ) . Additionally , whereas the repressed response to S . aureus consists of many transporters and cuticle components , the repressed response to P . aeruginosa is mostly represented by metabolic pathways ( Table S7 ) . Interestingly , both repressed responses included basement membrane genes ( N = 5 , p = 1 . 4E-6 for P . aeruginosa ) , suggesting host growth suppression by both types of infection . To investigate whether there were correlated or anti-correlated components of the responses to S . aureus and P . aeruginosa , we focused on genes whose expression changed both during infection with S . aureus and infection with P . aeruginosa . We broke down the two responses by plotting genes whose expression changed more than 2-fold with S . aureus ( Y axis in Figure 7B ) and with P . aeruginosa ( X axis in Figure 7B ) . Genes whose expression changed only during infection with one of the two pathogens were thus not included . This method defined four quadrants: I ) Genes induced by S . aureus and repressed by P . aeruginosa; II ) Genes induced by S . aureus and P . aeruginosa; III ) Genes repressed by S . aureus and induced by P . aeruginosa; and IV ) Genes repressed by S . aureus and P . aeruginosa ( Figure 7B ) . NextBio biogroup representation analysis ( see Materials and Methods ) on each quadrant failed to detect any over-represented biogroup in Quadrants I and III . However , in Quadrant II ( genes up-regulated by both pathogens ) several biogroups were over-represented , e . g . genes involved in detoxification , iron sequestration , and energy generation ( Table S2a , b ) . Likewise , biogroups over-represented in Quadrant IV ( down-regulated by both pathogens ) included transporters , cuticle components , and fatty acid ( FA ) β-oxidation ( Table S2a , b ) . The common repression of FA β-oxidation is not consistent with a starvation response , where β-oxidation is induced [56] . Furthermore , only 3 of 9 fasting repressed genes ( lbp-8 , acdh-1 , and F08A8 . 2 ) [56] were repressed during P . aeruginosa infection [44] . Similarly to S . aureus , the fasting-repressed gene hacd-1 was induced during P . aeruginosa infection [44] . Together , these observations show the distinct nature of the host response to distinct pathogens , and highlights metabolic components of early host responses to infection . Because overlapping gene expression changes measured by qRT-PCR do not necessarily imply the involvement of the same tissues during infection with different pathogens , we further investigated the expression of clec-60::GFP ( up-regulated by S . aureus and M . nematophilum , Figure S11A , B , C , D ) and F53A9 . 8::GFP ( up-regulated by S . aureus , M . nematophilum , and P . aeruginosa , Figure S11E , F , G , H ) . clec-60::GFP was induced by M . nematophilum and by S . aureus in the intestine ( Figure S11C , D ) , and down-regulated by P . aeruginosa ( Figure S11B ) compared with non-pathogenic E . coli ( Figure S11A ) . Likewise , F53A9 . 8::GFP was induced by all three pathogens in the intestine ( Figure S11F , G , H ) compared with E . coli controls ( Figure S11E ) . These data suggest that components of the host response are induced in the intestine by distinct pathogens . To determine whether S . aureus-induced genes have protective functions in host defense , we performed whole-animal RNAi knockdown of 42 of the 46 most highly up-regulated genes ( Table 1 ) , and identified 6 genes [tag-38 ( Glutamate decarboxylase/sphingosine phosphate lyase ) , sodh-1 ( sorbitol dehydrogenase ) , cyp-37B1 ( Cytochrome P450 ) , F43C11 . 7 ( F-box containing protein ) , math-38 ( MATH domain-containing signaling protein ) , and clec-70 ( secreted CTL ) ] whose decreased expression caused enhanced susceptibility to killing by S . aureus ( Figure 8A , B , C ) , but not P . aeruginosa ( Figure 8G ) . We also identified one gene , Y51H4A . 5 ( a putative intracellular lipase ) whose decreased expression caused slightly enhanced resistance to S . aureus killing , suggesting that its expression is detrimental to the host , or that it functions as a repressor of host defense ( Figure 8B ) . Importantly , the lifespan of animals continuously fed dsRNA-expressing , non-pathogenic E . coli was near wild type , except for F43C11 . 7 , which actually caused increased lifespan ( Figure 8D ) . These data show that S . aureus-induced genes have important functions in pathogen-specific host defense , and that the enhanced susceptibility to S . aureus mediated by RNAi knockdown is not due to a non-specific decrease in viability . Although the molecular identities of these genes offer clues to their potential functions , the elucidation of their exact mechanisms of action will require further study . Potential immune effectors identified in our analysis include antimicrobial peptides , lysozymes ( enzymes that degrade peptidoglycan in the bacterial cell wall ) , and CTLs [62] . To test whether lysozymes or CTLs induced by S . aureus can confer resistance to S . aureus-mediated killing when expressed to higher levels than in the wild type , we constructed transgenic C . elegans carrying multiple copies of lys-4 and lys-5 , clec-60 and clec-61 , or clec-70 and clec-71 ( each is a pair of genes that are adjacent to each other in the genome ) . Transgenic animals carrying the lys-4 , 5 or clec-60 , 61 clusters survived significantly longer than transgenic control animals ( Figure 8E; Figure S12 ) . Transgenic animals carrying the clec-70 , 71 cluster extrachromosomally did not exhibit enhanced resistance ( not shown ) ; however , three independent spontaneous integrant lines did ( Figure 8F ) . Interestingly , strains with multiple copies of either cluster of C-type lectin genes exhibited enhanced susceptibility to P . aeruginosa-mediated killing ( Figure 8H ) . Collectively , the data suggest that pathogen-induced intestinal expression of epithelial detoxifying and antimicrobial proteins is an important and pathogen-specific mechanism of C . elegans host defense against S . aureus infection .
Among the bacteria that cause intestinal infections in C . elegans , the best-studied is P . aeruginosa , a Gram-negative human pathogen . Despite many advances in understanding P . aeruginosa-C . elegans interactions , little was known about the morphological and cell biological consequences of infection in vivo . This report provides unprecedented high-resolution description of bacterial intestinal infections of clinical relevance in C . elegans , with emphasis on the comparative cytopathology of infection by P . aeruginosa and S . aureus . P . aeruginosa and S . aureus cause markedly different symptoms in C . elegans . During infection with P . aeruginosa , we observed marked intestinal distention , extracellular matrix accumulation in the intestinal lumen , extracellular material accumulation on the apical surface , enlargement of the rough endoplasmic reticulum ( RER ) , and abnormal autophagy in the host intestinal cells . The intestinal distention is likely a result of loss of cytoplasmic volume of the intestinal cells; the identity of the extracellular material is currently unknown . One possibility is that it is a biofilm matrix produced by P . aeruginosa , perhaps as a defensive mechanism . Alternatively , it could be produced by the host in response to P . aeruginosa , but not S . aureus . Further study is required to elucidate the origin of this substance . The cause of abnormal autophagy is also unknown; however , the aberrant structures we observed have also been described in unc-52 mutant animals , which are defective in the early steps of autophagosome assembly [63] . Thus , it is possible that P . aeruginosa infection causes autophagy arrest . It is tempting to speculate that this may represent a virulence mechanism deployed by P . aeruginosa to evade autophagic clearance of intracellular bacteria , an important host defense mechanism in the intestinal cells of C . elegans [64] and humans [65] . The intracellular invasion we observed provides a rationale for the benefit to P . aeruginosa of inhibiting host autophagy . gacA mutant P . aeruginosa was defective in inducing these phenotypes , indicating that GacA orchestrates virulence mechanisms related to host cell disruption in C . elegans . The molecular identity of these virulence mechanisms remains unknown; however , the observed putative OMVs may provide a mechanism for delivery of bacterial virulence factors to C . elegans intestinal cells . Perhaps the reliance on OMVs for virulence may explain why P . aeruginosa defective in type III secretion , an alternative mechanism of virulence factor delivery to host cells , are not defective in nematode killing [66] . OMV production is induced in P . aeruginosa by cellular stress [67]; therefore , the abundant OMV production we observed may indicate that P . aeruginosa perceives the intestinal lumen as a stressful environment , presumably as a result of defensive host factors secreted by the intestinal cells . The molecular identity of such putative intestinal defense factors remains unknown . In previous work , we defined the early transcriptional host response to P . aeruginosa [68] . Among the genes that are up-regulated by P . aeruginosa infection , several encode putative antimicrobial factors such as ShK toxins . Whatever the molecular identity of the antibacterial factors induced during P . aeruginosa infection , heat-killed P . aeruginosa did not induce a set of 10 biomarkers of the response . This observation is consistent with our previous studies suggesting that P . aeruginosa virulence may be required , at least partially , for response induction [44] , [69] . These results are consistent with several interpretations . It is possible that detection of virulent P . aeruginosa is mediated by recognition of the damage inflicted on the host cells , e . g . , via DAMP perception [70] , or by recognition of PAMPs in the context of host damage , in what has come to be known as a pattern of pathogenesis [71] . Alternatively , PAMPs released only by live bacteria ( PAMP-per vitae , or PAMP-PV ) may be specifically recognized by C . elegans as a trigger for the response [71] . A third option is the release of PAMP-post mortem ( PAMP-PM ) upon heat inactivation of P . aeruginosa , which in turn could dampen the host response . PAMP-PM detection has been proposed to be a mechanism used by mammals to limit inflammatory damage to the host once the infection has been controlled [71] . Finally , detection of P . aeruginosa could be mediated by heat-labile signals that were destroyed during heat-inactivation in our experiments . Although further work is required to conclusively show which of these scenarios is correct , we have found that P . aeruginosa strains with increasing levels of virulence cause increasing levels of induction of C . elegans host response gene irg-1 [69] , suggesting that the extent of host cell damage determines the magnitude of the host response to the hypothetical P . aeruginosa-produced PAMPs or DAMPs that are detected . During S . aureus infection , we observed rapid intestinal colonization , swelling of the anus , and effacement and destruction of intestinal epithelial cells , providing important mechanistic information about S . aureus-mediated pathogenesis of host epithelial cells in vivo . The molecular mechanism of cell destruction remains unknown; here we show that it is hemolysin-independent and is abrogated by heat inactivation of S . aureus . One explanation for this observation is that host tissue damage may be caused by the active pathogen , and not an unbridled host response . Alternatively , unknown heat-labile S . aureus toxins may cause cell lysis , as can occur in human cells [72]; further study is required to distinguish between these possibilities . Previous experiments using fertile animals showed that α-hemolysin Δhla mutant S . aureus were defective in C . elegans killing [17] . To our surprise , the Δhla Δhlb , Δhla Δhlg , or Δhla Δhlb Δhlg hemolysin mutants did not exhibit any killing defect in our assay using sterile animals ( Figure S7 ) . Thus , it is possible that α-hemolysin-mediated killing of C . elegans requires internal hatching of eggs retained inside the mother as a result of stress . Eventually , the pathogenic process results in internal tissue degradation and nematode death . These steps recapitulate key features of S . aureus infection in mammals both in vivo and in vitro , e . g . enterocyte effacement during intestinal infection , and cell lysis [37] , [38] , [39] , [40] , [41] , [42] , [43] . Thus , we propose that the C . elegans-S . aureus model has significant relevance to the study of conserved virulence mechanisms used by S . aureus to evade host epithelial defenses and attack host epithelial cells in general . Transcriptional profiling showed that S . aureus infection elicits changes in expression of a minor fraction of the total genome of ∼22 , 000 genes , indicating high specificity . This early response does not require live bacteria , suggesting that it involves detection of S . aureus per se ( perhaps through PAMP perception ) as opposed to indirectly through host cell damage ( through DAMPs ) [70] . Whatever the relevant PAMPs are , it is clear that they are not shared between S . aureus and B . subtilis , also a Gram-positive bacterium . For example , the peptidoglycan differs greatly between these two species [73] , and could potentially be differentially sensed by C . elegans . Other possibilities include differential detection of surface-expressed lipoteichoic acids or differentially expressed surface proteins . Further work is required to elucidate the nature of such signal ( s ) . Our results provide insight into the cellular biological effects of pathogenic infection on the epithelial barrier in vivo , as well as the early defense mechanisms deployed by C . elegans to fend off attack . The affected genes represent evolutionarily conserved categories relevant to human innate immunity . To gain insight into evolutionarily conserved effector mechanisms of host defense , we compared the genes up-regulated in C . elegans with previously published data using human neutrophils , which are important effector cells in human innate immunity [74] . When grouped by molecular function , we found some overlapping functional classes during the C . elegans and human neutrophil responses to S . aureus infection ( Table S4 ) . These classes included detoxification factors [e . g . transporters , UDP-glucuronosyltransferases ( UGTs ) , cytochrome P450s , GSTs , and flavin-containing monooxygenases ( FMOs ) , [75]] , antimicrobial effectors ( e . g . CTLs , peptidases , and proteases ) , galectins [76] , and signaling components including EGF-like domain containing proteins , Cdc42 guanyl nucleotide exchange factors ( GEFs ) , F-box proteins , mitogen-activated protein kinases ( MAPKs ) , and Leucine-rich repeat domain ( LRR ) containing proteins . This observation shows that the C . elegans host response to infection shares important components with the human cellular host response , and suggests that human innate responses have ancient components that are conserved across phylogeny . Indeed , it is thought that modern vertebrate innate immunity represents an accretion of ancient invertebrate innate defenses [77] . Comparative genomics identified pathogen-specific as well as pathogen-shared components of the host response . This observation , consistent with similar ones recently made by others using different approaches [50] , [78] , [79] , illustrates how diverse pathogens affect distinct aspects of host physiology as reflected in the distinct nature of the host responses . We also found overlap in gene expression patterns among the responses to three different pathogens , S . aureus ( intestinal infection , Gram-positive ) , M . nematophilum ( cuticular infection , Gram-positive ) , and P . aeruginosa ( intestinal infection , Gram-negative ) , defining a core induced response that involves intracellular detoxification , iron sequestration , and sugar binding . In addition to a common set of up-regulated genes , we observed a repressed core response common to S . aureus and P . aeruginosa that involves anion transport , growth-related genes , lipid- and alcohol-metabolic genes , and acyl-CoA dehydrogenases . The fact that metabolic regulation is a major component of the C . elegans host response to bacterial pathogens provides a rationale to investigate metabolic changes that occur in higher organisms as a result of infection , particularly in innate defense tissues such as epithelia . Recent reports suggest a significant metabolic component during the host response in mammals as well [80] , [81] , [82] . The non-overlapping responses to S . aureus and P . aeruginosa may reflect the different virulence strategies of the two pathogens and/or may be a consequence of the distinct molecular composition of Gram-positive and -negative cell walls . Further studies are required to dissect the relative contribution of each factor , including the survey of additional Gram-positive and -negative infections in C . elegans . In a first step along those lines , we found that the Gram-positive non-pathogenic bacterium B . subtilis does not induce the same set of 10 biomarkers as S . aureus . Additionally , of 531 genes up-regulated by the Gram-positive pathogen E . faecalis [50] , only 15 were shared with the response to S . aureus ( Table S5a ) . The same report characterized the C . elegans late host response ( i . e . , 24 h ) to four additional pathogens , identifying a set of shared genes that defined a pathogen-shared necrotic response [50] . During S . aureus infection we observed up-regulation of none of 16 up-regulated shared late response genes , and down-regulation of only three of six down-regulated shared response genes ( Table S5c ) , suggesting that the host response evolves significantly over time . Mammalian intestinal epithelial cells directly sense and respond to bacterial stimulation [2] , by inducing the expression of antimicrobial genes such as CTLs [53] . Similarly , most of the early C . elegans response to S . aureus occurs in the epithelial cells of the intestine . In Drosophila , mice , and humans , TLRs are important receptors that drive the activation of signaling cascades downstream of microbial stimulation . In C . elegans , however , loss of function of the sole TLR does not result in a defective immune response to S . aureus . Furthermore , C . elegans does not have an NF-kB homolog nor an inflammasome , raising the possibility that in mammals as in C . elegans , at least a portion of the immune response to S . aureus may be regulated independently of the TLR/NF-kB signaling axis [6] , [51] , [83] . Indeed , we previously reported that β-catenin and HOX genes are required for perception of pathogenic attack by S . aureus to drive the expression of epithelial host response genes [7] . Significantly , we also found that β-catenin and HOX proteins modulated NF-κB signaling in a human epithelial cell line during TLR2 stimulation , illustrating that previously unknown human innate immunity pathways can be identified using C . elegans [7] . We identified 6 host factors out of 42 tested whose lowered expression caused enhanced susceptibility to S . aureus . This corresponds to a 14% hit rate , which was much greater than expected; we had assumed that there would be significant functional redundancy among C . elegans immune effectors . Moreover , RNAi typically exhibits incomplete penetrance and expressivity . In addition to these 6 genes , knockdown of Y51H4A . 5 ( lipase ) caused mild resistance to S . aureus mediated killing , suggesting that Y51H4A . 5 acts to limit survival as a negative regulator of the host response or by harming the host instead of protecting it . RNAi of F43C11 . 7 caused enhanced susceptibility to S . aureus , yet extended lifespan on non-pathogenic E . coli . This is an example of genes that have opposite effects on host defense and lifespan regulation , indicating that these related processes are genetically separable [84] . Increased expression of host defense genes provides a mechanistic explanation of C . elegans defenses , as we found that animals carrying multiple copies of three genomic clusters of lysozymes or CTLs exhibited enhanced resistance to S . aureus . Lysozymes are well-known , evolutionarily ancient antibacterial effector molecules that degrade peptidoglycan and are also produced in human intestinal epithelial cells [1] . Recent studies have shown that some vertebrate CTLs , including human HIP/PAP , have direct bactericidal activity [53] . clec-70 and clec-71 share similar domain architectures with HIP , and clec-60 and clec-61 share similarities with arthropod receptor CTLs , suggesting that they may function respectively as antimicrobials or receptors in the C . elegans host response [85] , [86] , [87] . Collectively , these observations show that the large number of pathogen-response genes contribute cumulative , incremental defense functions to host survival . It is interesting that elements of the C . elegans immune response enhance host survival during infection with S . aureus ( a human pathogen ) , supporting the notion that pathogen detection and response , as well as mechanisms of bacterial pathogenesis , share conserved features among distantly related hosts or microbes , respectively . It has been proposed that pathogens have experienced stepwise additions of virulence factors , as they evolved to survive different host antimicrobial responses , and to colonize new niches [12] . Our studies of the C . elegans-S . aureus system may thus probe an early step in the evolution of S . aureus as a pathogen and its interaction with prototypical metazoan epithelial cells . In humans , unknown host and bacterial factors determine whether S . aureus will become an innocuous member of the normal microbiota , or whether it will switch to a more virulent state and become a serious pathogen [23] . In this light , studies of the C . elegans intestinal epithelial response to S . aureus provide a unique starting point to identify previously unknown signaling pathways and molecular mechanisms of host immune response to bacterial virulence . Understanding how S . aureus disrupts host defense and causes host damage and death is critical to identifying new therapeutic targets to treat infectious disease .
C . elegans was grown on nematode-growth media ( NGM ) plates seeded with E . coli OP50-1 at 15–20°C according to standard procedures [88] . C . elegans strains used in this study are detailed in Table S6a . Bacterial strains are detailed in Table S6b . Wild type N2 Bristol animals were synchronized by hypochlorite treatment and L1 arrest and incubated on NGM plates seeded with E . coli OP50-1 . Late L4 animals were collected and plated on 15 cm TSA plates seeded with live S . aureus NCTC8325 or heat-killed NCTC8325 , and parallel NGM plates seeded with OP50-1 . After 12 , 24 , and 36 h incubation at 25°C , animals were collected and incubated in fixation buffer ( 2 . 5% glutaraldehyde , 1 . 0% paraformaldehyde in 0 . 05 M sodium cacodylate buffer , pH 7 . 4 plus 3 . 0% sucrose ) . During the initiation of fixation , animals were cut in half with a surgical blade in a drop of fixative under a dissecting microscope , fixed overnight at 4°C , rinsed in 0 . 1 M cacodylate buffer , post-fixed in 1 . 0% osmium tetroxide 0 . 1 M cacodylate buffer , rinsed in buffer and water , and stained en bloc in 2% aqueous uranyl acetate . After rinsing in water , animals were embedded in 2% agarose in phosphate buffer saline , dehydrated through a graded series of ethanol washes to 100% , then 100% propylene oxide , and finally 1∶1 propylene oxide:EPON overnight . Blocks were infiltrated in 100% EPON and then embedded in fresh EPON overnight at 60°C . Thin sections were cut on a Reichert Ultracut E ultramicrotome and collected on formvar-coated gold grids . Sections were post-stained with uranyl acetate and lead citrate and viewed using a JEOL 1011 transmission electron microscope at 80 kV with an AMT digital imaging system ( Advanced Microscopy Techniques , Danvers , MA ) . For each observation , whenever possible at least 10 cross-sections were evaluated , and representative images were chosen . All assays were conducted at 25°C , 65% relative humidity . Animals were scored as alive or dead by gentle prodding with a platinum wire . Kaplan-Meier statistical analyses were performed using the software Prism ( GraphPad , http://www . graphpad . com ) . Survival data were compared as described using the log-rank test . Data are represented as median survival ( MS ) or lethal time – 50 ( LT50 ) when MS values were skewed by small number of timepoints , N ( number of deaths/censored ) , and p value . A p-value <0 . 05 was considered significantly different from control . Animals were treated essentially as described for killing assays described above , with the following modifications . For S . aureus infection assays , infected samples were compared to parallel samples feeding on E . coli OP50 , heat-killed by 30 min incubation at 95°C , plated on the same TSA medium . All strains compared were grown in parallel . Total RNA was extracted using TRI Reagent , and reverse transcribed using the Superscript III kit ( Invitrogen ) . cDNA was subjected to qRT-PCR analysis using SYBR green detection ( BIO-RAD SYBR Green supermix ) on iCycler ( Bio-Rad , http://www . bio-rad . com ) and RealPlus ( Eppendorf , Germany ) machines . Primers for qRT-PCR were designed using Primer3Plus ( Massachusetts Institute of Technology , http://www . bioinformatics . nl/cgi-bin/primer3plus/primer3plus . cgi ) , checked for specificity against the C . elegans genome and tested for efficiency with a dilution series of template . Primer sequences are available upon request . All Ct values are normalized against the control gene snb-1 , which did not vary under conditions being tested . Fold change was calculated using the Pfaffl method [93] . We found some variability in gene induction levels from experiment to experiment . The source of this variation has not been conclusively ascertained; however , we suspect it may derive from differences between batches of agar plates used for infection assays . Importantly , all experiments were repeated at least twice ( biological replicates ) and were internally controlled . Additionally , despite numerical variability in fold induction , all results were internally consistent . PCR primers to amplify 1665 bp of sequence upstream of the clec-60 start site , 3505 bp upstream of the clec-70 start site , 1774 bp upstream of the fmo-2 start site , and 913 bp upstream of the ilys-3 start site were designed using the online PCR primer design tool provided by the British Columbia Genome Sciences Center ( http://elegans . bcgsc . bc . ca/promoter_primers/index . html ) . Splicing by overlapping extension PCR ( SOE-PCR ) was used as described [94] to generate promoter-GFP fusion PCR fragments , which were transformed at 3 ng/ µl into wild type animals by microinjection with 40 ng/ µl of a myo-2::NLS::mCherry construct as coinjection marker used to identify transgenic animals ( courtesy of J . Kaplan , Massachusetts General Hospital ) . Primer sequences are available upon request . L4 animals carrying extrachromosomal arrays were transferred from NGM plates seeded with OP50-1 to S . aureus , P . aeruginosa , or M . nematophilum killing plates essentially as described above . After incubation , animals were mounted on glass slides with 2% agarose pads , anesthetized with 30 mM NaN3 , and immediately used for imaging . Exposure times were set for the most highly expressed condition and kept constant throughout each experiment . Gene expression microarray data files of S . aureus infection were obtained from Gene Expression Omnibus ( GEO accession: GSE2405 ) and analyzed . The samples were derived from human polymorphonuclear leukocytes ( PMNs ) from three healthy donors using a separate HU133A GeneChip ( Affymetrix ) for each donor [74] . We examined the data from PMNs that were either uninfected or infected with live S . aureus for 9 hours . The dataset was MAS5 . 0-normalized and filtered by excluding probe sets with 100% ‘absent’ calls ( MAS5 . 0 algorithm ) across all samples . FDR analysis ( q-value<0 . 005 ) with 1000 permutations using significance analysis of microarrays [98] was performed to identify genes that were differentially induced in S . aureus-infected PMNs versus uninfected controls . Images were acquired using a Zeiss AXIO Imager Z1 microscope with an Zeiss AxioCam HRm camera and Axiovision 4 . 6 ( Zeiss ) software . Image cropping and minimal manipulation were performed using Photoshop ( Adobe ) . Gene Public Name , Gene WormBase ID , Source GenBank ID , Gene CGC Name; bar-1 , WBGene00000238 , U46673 , bar-1; col-63 , WBGene00000639 , Z81143 , col-63; col-98 , WBGene00000673 , Z81503 , col-98; cpr-2 , WBGene00000782 , Z81531 , cpr-2; egl-5 , WBGene00001174 , L15201 , egl-5; exc-5 , WBGene00001366 , Z68159 , exc-5; fmo-2 , WBGene00001477 , Z70286 , fmo-2; ins-11 , WBGene00002094 , U41279 , ins-11; lys-2 , WBGene00003091 , AL021479 , lys-2; lys-5 , WBGene00003094 , Z73427 , lys-5; mpk-1 , WBGene00003401 , Z46937 , mpk-1; pmk-1 , WBGene00004055 , U58752 , pmk-1; sod-3 , WBGene00004932 , U42844 , sod-3; tol-1 , WBGene00006593 , AF348166 , tol-1; unc-32 , WBGene00006768 , Z11115 , unc-32; C32H11 . 1 , WBGene00007864 , NM_070062 . 2; C50F4 . 9 , WBGene00008234 , Z70750; F01D5 . 2 , WBGene00008493 , Z81493; acs-2 , WBGene00009221 , Z81071 , acs-2; F55G11 . 2 , WBGene00010123 , Z82272; clec-60 , WBGene00014046 , Z49132 , clec-60; clec-61 , WBGene00014047 , Z49132 , clec-61; clec-52 , WBGene00015052 , U58752 , clec-52; C23G10 . 6 , WBGene00016013 , U39851; C30G12 . 2 , WBGene00016274 , U21319; ilys-3 , WBGene00016670 , AF067611 , ilys-3; C49G7 . 5 , WBGene00016783 , AF016418; acdh-1 , WBGene00016943 , AC006625 , acdh-1; F49F1 . 6 , WBGene00018646 , AF100656; F53A9 . 8 , WBGene00018731 , U23523; F53E10 . 4 , WBGene00018760 , U88177; clec-70 , WBGene00021581 , AC024785 , clec-70; clec-71 , WBGene00021582 , AC024785 , clec-71; Y65B4BR . 1 , WBGene00022040 , AC024847 . | Pseudomonas aeruginosa and Staphylococcus aureus are bacteria that can form part of the human microbiota , but can also cause severe disease . Despite their great clinical importance , little is known about their interactions with the human host before disease onset , particularly regarding the molecules that host cells use to prevent and combat infection . The invertebrate Caenorhabditis elegans is a powerful model host to study host-pathogen interactions and , because of its simplicity and highly-developed experimental methods , can illuminate fundamental mechanisms of bacterial pathogenesis . We used C . elegans to understand the cellular events that occur during early stages of P . aeruginosa and S . aureus infection . We found that P . aeruginosa slowly colonizes the intestine , producing virulence-related membrane vesicles , and causing accumulation of electron-dense biofilm-like material on intestinal cells and abnormalities in them . In contrast , S . aureus colonizes rapidly , disrupting microvilli and lysing host cells . We found that these different strategies result in different host gene expression responses . Focusing on S . aureus infection , the C . elegans response is mainly microbicidal and detoxifying , aiding host survival and bearing similarities with the human response . Our study provides new insights into mechanisms that P . aeruginosa and S . aureus use to cause disease , and into C . elegans defenses , with potential implications for human immunity and disease . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"microbiology/innate",
"immunity",
"immunology/innate",
"immunity",
"infectious",
"diseases/bacterial",
"infections"
] | 2010 | Distinct Pathogenesis and Host Responses during Infection of C. elegans by P. aeruginosa and S. aureus |
The implementation of vector control interventions and potential introduction new tools requires baseline data to evaluate their direct and indirect effects . The objective of the study is to present the seroprevalence of dengue infection in a cohort of children 0 to 15 years old followed during 2015 to 2016 , the risk factors and the role of enhanced surveillance strategies in three urban sites ( Merida , Ticul and Progreso ) in Yucatan , Mexico . A cohort of school children and their family members was randomly selected in three urban areas with different demographic , social conditions and levels of transmission . We included results from 1 , 844 children aged 0 to 15 years . Serum samples were tested for IgG , NS1 and IgM . Enhanced surveillance strategies were established in schools ( absenteeism ) and cohort families ( toll-free number ) . Seroprevalence in children 0 to 15 years old was 46 . 8 ( CI 95% 44 . 1–49 . 6 ) with no difference by sex except in Ticul . Prevalence increased with age and was significantly lower in 0 to 5 years old ( 26 . 9% , 95% CI:18 . 4–35 . 4 ) compared with 6 to 8 years old ( 43 . 9% , 95% CI:40 . 1–47 . 7 ) and 9 to 15 years old ( 61 . 4% , 95% CI:58 . 0–64 . 8 ) . Sharing the domestic space with other families increased the risk 1 . 7 times over the individual families that own or rented their house , while risk was significantly higher when kitchen and bathroom were outside . Complete protection with screens in doors and windows decreased risk of infection . Seroprevalence was significantly higher in the medium and high risk areas . The prevalence of antibodies in children 0 to 15 years in three urban settings in the state of Yucatan describe the high exposure and the heterogenous transmission of dengue virus by risk areas and between schools in the study sites . The enhanced surveillance strategy was useful to improve detection of dengue cases with the coincident transmission of chikungunya and Zika viruses .
Dengue is a major public health problem in Latin America due to the increasing trend of cases , the vast urban areas affected , and the complexity of controlling a vector that has adapted to human dwellings in tropical and subtropical urban contexts [1] . Accurate estimates of the burden of dengue [2] are difficult because of the high proportion of asymptomatic infections , the syndromic nature of the clinical spectrum that allows for misdiagnosis with other viral infections [3] , the limited capacities of the surveillance systems , and the low demand for health services by affected populations [4–6] . Transmission of the four dengue serotypes in endemic countries is heterogeneous with respect to the age groups affected , the seasonality , and the intensity and severity of epidemics[7] . An improved understanding of the complex dynamic of factors involved in dengue transmission requires the characterization of different parameters related to the incidence of asymptomatic , sub-clinical and symptomatic infections [8 , 9]; the prevalence and seroconversion rates by age group and sex; the herd immunity to specific serotypes [10]; the profile of primary and secondary infections and risk factors associated with severe dengue; as well as their relationship with the entomological variables at the individual , household , neighborhood , locality and regional levels [11–14] . Prospective studies have become crucial for understanding dengue transmission in urban settings and are invaluable in providing the data required to effectively evaluate the impact of traditional and innovative control strategies [15 , 16] . In endemic areas , transmission dynamics can be better understood with the longitudinal study of young and naïve populations [17] . Selecting school children as the basis for a cohort has several advantages . They are generally susceptible to dengue infection; can be involved and followed-up for longer periods through their attendance to nearby schools; their families are responsive and supportive to health initiatives arising from the educational institutions , and their households are near the school facilitating logistics for follow-up and allowing for a febrile and absentee school-based surveillance system [18] to monitor under reporting of febrile and dengue cases . In addition , there are concerns about the specific risks that school grounds may have in triggering transmission in certain environments [19] . The availability of a licensed vaccine poses different challenges to current dengue control programs since its gradual introduction is not expected to cover all susceptible and at-risk populations or even provide complete immunity in target groups . In each of these populations there are several questions that need to be addressed regarding the clinical spectrum and transmission risks where the vaccine or any other control innovation may provide a potential benefit [20] . The objective was to describe the seroprevalence of dengue infection in a cohort of children from elementary and middle schools in three urban sites with different socio-demographic-economic profiles and transmission patterns in the state of Yucatan , Mexico . This study presents the seroprevalence status , the socio-demographic risk factors associated with dengue infection , and the results of enhanced surveillance strategies established to support detection of dengue cases from 2015 through 2016 .
Merida is the capital and major human settlement of Yucatan State , with 892 , 363 inhabitants ( 42 . 5% of the state population ) and 257 , 826 ( 46 . 4% ) houses . It is the most important economic city , concentrating 50% of the industrial activity . Approximately 1 million national and 250 , 000 international tourists visit Merida every year [21] . Climate in Yucatan is warm and humid , and the rainy season falls between June and October; the mean annual temperature is 25 . 9◦C ( 19 . 5 to 33 . 6 ) and annual precipitation is 1050 ( mm ) . Merida concentrates ~60% of all dengue cases reported in the state . Ticul is an urban locality , located 96 kilometers south of Merida with 40 , 161 people and about 9 , 808 houses , concentrates around 3% of all dengue cases in the state . Progreso is the major seaport located 27 km north of Merida . It has 59 , 122 people and about 16 , 020 houses . Progreso is the most popular beach resort and tourist destination for many local citizens as well as national and international visitors ( 291 , 709 tourists and 136 cruises ship every year ) . Consequently , most inhabitants of Progreso ( 60 . 4% ) are involved in commercial and tourist services . Progreso represents around 1% of all dengue cases reported in Yucatan . The cohort study was initially defined by a random selection of five extensive geographical areas within those cities . These included: two low risk areas ( one urban area in the north of Merida and the town of Progreso ) ; two medium risk areas ( one urban area in central Merida and the town of Ticul ) ; and one high-risk urban area in the south of Merida . The definition of risk was determined by the historical reports of the number of dengue cases , the percent of cases reported every year and the continuous transmission during 6 to 8 or more weeks every year provided by the state surveillance system [22] . The entire study population was composed of a cohort of children ( index children ) in elementary and middle schools together with their family members sharing the same home with the index children ( family equals number of index children ) . This report describes the results from the cohort of index school children and their siblings up to 15 years of age . The State Ministry of Education of Yucatan provided a list of elementary public schools located in the selected areas , and a cohort of children from 1st to 3rd grade ( 6 to 8 years old ) was randomly selected from eight schools in Merida , four in Progreso and two from Ticul . A convenience sample of 50 children per grade was defined to gather 150 school children ( index children ) in each of the five risk areas ( 450 for Merida ( 150 in each low , middle high risk areas ) , 150 for Ticul ( middle risk ) and 150 for Progreso ( low risk ) ) . Because children from different grades could come from the same family , we restricted the selection to one child per family to have 50 different and independent families per grade . Recruitment of school-aged children ( between 6 to 12 years ) included invitation to all other members of the household of the children enrolled . Consent and assent forms were obtained individually from each adult and from parents in the case of children and participants younger than 18 years old and were signed before blood samples were taken . Exclusion criteria included refusal to participate or plans to move outside of the study area during the months following enrolment . The enrolment of new school children in the 2nd year ( 2016 ) was designed to incorporate 50 new 1st grade children per risk area from the same elementary schools or additional ones when required . Based on the results of the dengue vaccine trials [23] the target groups for the vaccine were children over 9 years old instead of under-five years old . Therefore , the recruitment scheme changed to include more children aged 9 to 12 years old from new middle schools . From the additional 150 new children required from middle schools in Merida , only 134 were recruited along with 37 from elementary schools ( 171 new index children ) . Ticul and Progreso recruited 50 middle school children each plus 19 and 16 new index children from elementary schools , respectively . Baseline and follow-up evaluations of the cohort population were obtained during the period December 2014 through August 2016 with baseline demographic information , clinical history of dengue and blood samples taken for serological evidence of dengue virus ( DENV ) infection after the annual transmission season ( August to December ) . Baseline and first follow up evaluation are presented here . The data collected included individual , family and household questionnaires and was obtained by a multidisciplinary field team called “Familias sin Dengue” ( FSD = Families without dengue ) integrated by physicians , nurses , social scientists and technical personnel . Basic data regarding house characteristics included construction material , number of rooms , sanitation services ( potable water , sewage , garbage collection ) , water use patterns , and physical protection of windows such as screens in doors and windows . The individual and family questionnaires included basic demographic data ( age , sex , education level , occupation , among others ) together with a clinical history of dengue including signs and symptoms , dates of occurrence , access or utilization of health services and hospitalization . The febrile questionnaire included data regarding febrile episodes , symptoms , dates , duration , severity , movements outside the area , utilization of health services , contacts and blood sample results ( serology ) . Symptomatic dengue was detected through passive surveillance for dengue-associated symptoms and absenteeism from schools . The clinical team of FSD visited schools every week , checked attendance , and visited the homes of absentees . If absentees had a febrile illness in the previous week , they were evaluated with a physical exam and an acute blood sample to confirm infection . In addition , this population was monitored for febrile symptoms and suspected dengue illness through a toll-free number where parents could call for medical assistance when febrile disease appeared in any member of the family . This free of charge telephone line was directed only to the members of the cohort so as to voluntarily report febrile cases to the team ( FSD ) where a physician in charge would respond to their request . Participants were defined as lost to follow up after a full year had passed since their previous blood sample , despite repeated attempts to locate the participant , or if there was a verifiable reason for dropping from the study ( voluntary request from the participant , movement from the study area ) . Baseline serum samples were taken to test for IgG antibodies by capture enzyme-linked immunosorbent assay ( ELISA-Panbio ) . A 5ml of peripheral venous blood was obtained with BD Vacutainer and centrifuged at 3000 rpm ( Bio-Lion XC-L4 ) . The serum was stored at 4±2°C and aliquots obtained with Scilogex micropette plus autoclavable pipettor of 500μl at -70°C ( Eppendorf CryoCube F570-86 Upright ) . Following the reference values , negative , equivocal and positive results were determined as <9 , 9–11 and >11 Panbio units , respectively . Febrile episodes in enhanced surveillance were classified as DENV infections based on NS1 and IgM serology . Descriptive analysis of the cohort included school children and their siblings aged 0 to 15 years . Participants were stratified in three age groups ( 0 to 5 , 6 to 8 and 9 to 15 years old ) . Dengue IgM and IgG results above >11 Panbio units were considered positive for dengue and were included as the dependent variable in the analysis . The logistic regression analysis was adjusted by age and sex and was used to identify risk factors for dengue infection . Odds ratios ( ORs ) and their 95% confidence intervals were calculated and statistically significant differences ( p<0 . 05 ) were included in the final model . The analysis was done with STATA 14 . 2 . The protocol was approved by the Ethics and Research Committee from the O´Horan General Hospital from the state Ministry of Health , Register No . CEI-0-34-1-14 and the review board at the Fred Hutchinson Cancer Research Center .
Sex distribution in the cohort of school children was very similar ( 48 . 7% women and 51 . 3% men ) in all risk areas although low risk Merida had 59 . 1% male children . The age distribution of all children was 14 . 3% for children under five years old , 39 . 9% for 6 to 8 years old and 45 . 9% for children aged 9 to 15 years old . Ticul had the lowest proportion of children under five years ( 8 . 6% ) ( Table 2 ) . House property patterns in Yucatan showed that most of the houses were privately owned ( 70 . 5% ) , few were rented ( 4 . 5% ) and 25% shared with other families ( usually kin ) . Construction materials of walls , floor and roof were cement ( >95% ) although palm roofs in Progreso ( 7 . 7% ) and Ticul ( 7 . 9% ) described particular socioeconomic conditions of some families participating in the study . Most of the houses ( 72% ) had 2 to 3 rooms while 18 . 8% of houses in Ticul , low risk Merida ( 18 . 2% ) and high risk Merida ( 15 . 9% ) had 4 to 5 rooms . Protection with screens in doors and windows was a common practice in this region with almost 95% of houses with at least one door or window protected . The presence of kitchen and bathroom outside the house was much more common in Ticul ( 25 . 7% and 24 . 1% respectively ) than in Progreso ( 3 . 3% and 12 . 6% ) and the high risk area in Merida ( 7 . 7% and 12 . 6% ) . Access to potable water was also a common feature ( >90% ) in the three cities but the need to store water for several domestic activities was also very common ( 45 . 9% ) in all sites , particularly in Ticul that reported a higher need to store potable water ( 78% ) . Garbage collection was a widespread public service in the areas , although in Ticul , 19 . 4% of households reported the need to burn or throw away the garbage ( Table 3 ) . The percentage of blood sampling in 1 , 521 children in the 2015 cohort and 1 , 844 children in 2016 was 85% ( Fig 1 ) . Ticul had the highest coverage ( 90 . 5% ) of blood samples , Merida had similar coverages by risk areas ( 87% ) , while Progreso presented the lowest coverage ( 75 . 6% ) . There were 888 ( 68 . 3% ) paired samples in children . Again , Ticul led the blood sampling with 79 . 3% , followed by Progreso ( 70% ) and Mérida ( 63 . 7% ) . ( S1 Table ) . Seroprevalence of 0 to 15 years old was 46 . 8 ( CI 95% 44 . 1–49 . 6 ) . Overall seroprevalence to dengue showed no difference in females ( 53 . 9% , 95% CI: 50 . 4–57 . 4 ) compared to males ( 49 . 4% , 95% CI: 45 . 8–52 . 9 ) and differences by age group or city were not significant except in Ticul ( 61% , 95% CI: 53 . 9–67 . 9 ) for female vs male ( 49 . 4% 95% CI: 45 . 82–52 . 87 ) . Seroprevalence increased with age and was significantly lower in 0 to 5 years old ( 26 . 9% , 95% CI:18 . 4–35 . 4 ) compared with children 6 to 8 years old ( 43 . 9% , 95% CI:40 . 1–47 . 7 ) and 9 to 15 years old ( 61 . 4% , 95% CI:58 . 0–64 . 8 ) . Seroprevalence by age group and risk area showed significant differences between the 0 to 5 years old compared to the 9 to 15 years old in low risk Merida ( 15 . 8% , 95% CI: -0 . 61–32 . 2 vs 61 . 5% , 95% CI:53 . 2–69 . 9 ) , medium risk Merida ( 13 . 3% , 95% CI:1 . 2–25 . 5 vs 66 . 2% , 95% CI:58 . 9–73 . 6 ) and Ticul ( 30% , 95% CI:1 . 6–58 . 4 vs 69 . 0% , 95% CI:62 . 3–75 . 6 ) but not in Progreso or high risk Merida . Differences between 6 to 8 years old and the 9 to 15 age groups were significant in all risk areas except Progreso ( Fig 2 ) . Seroprevalence by school showed variations in dengue exposure within areas of risks . The lowest rate in low risk areas like Progreso or low risk area in Merida are not different from the lowest rate in Ticul or high risk Merida . Similar patterns appeared in the schools with the highest seroprevalence in each risk area ( Fig 3 ) . The multivariate analysis of children with blood results showed no difference by sex but dengue risk increased with age , being four times higher for those 9 to 15 years old compared to children under five years old . While parents reported very few cases of dengue in the previous year , the history of dengue was significant as well as the report of dengue confirmation by health personnel . Regarding the risks derived from household conditions obtained by the questionnaires: sharing the domestic space with other families increased the risk 1 . 7 times over the individual families that own or rented their house , and the risk of dengue was significantly higher when kitchen and bathroom were located outside the house . Protection with screens of windows showed a good level of protection only when screens cover all the windows . While bivariate analysis of the prevalence of dengue by urban areas showed discrete differences , the analysis combining areas of risk ( Progreso and low risk Merida vs . medium risk Merida and Ticul , and high risk Merida alone ) pointed out that risk of infection was significantly higher in the medium and high risk areas compared to those in the low level risk ( Table 4 ) . Surveillance of absenteeism in the schools showed 98 cases of absentees due to fever ( January to June 2015 ) . An outbreak of 57 cases of chickenpox was detected along with three likely cases of dengue that had negative test results . During the second semester , an outbreak of chikungunya emerged in the State of Yucatan , four schools closed for four days and 20 probable cases of dengue and 34 of chikungunya were detected . In Ticul , 44 children were detected as absentees and only three probable cases of dengue were detected with negative results . A total of seven cases of chikungunya were reported . In Progreso , 11 children were absent , six cases were respiratory infections , two diarrheas , two cases of conjunctivitis and one with chickenpox . From September to November , three fever cases were reported , two suspected cases of dengue that were negative . During 2016 , absenteeism surveillance reported 434 total cases of absenteeism , 219 cases ( 50% ) due to fever . Four cases were surgical pathologies and only two clinical dengue cases; 107 cases of respiratory diseases and 78 chickenpox cases ( 46% and 51% were in Merida and Ticul schools ) , 28 cases were not located . A total of 53 cases of absenteeism were reported in cohort students: 23% ( 12/53 ) were non-febrile cases; 77% due to febrile pathologies such as allergy symptoms ( 1 ) , abdominal pain ( 1 ) , respiratory tract infections ( 27 ) , urinary infection ( 1 ) , chickenpox ( 10 ) and one family was not located . During 2015 , 373 telephone calls were received by the FSD group , 32 . 5% ( n = 121 ) were not related to a febrile episode and patients were referred to their family physician or clinic . Around 59 . 1% ( n = 149/252 ) of febrile patients that contact the dengue line had previously consulted a physician who gave the diagnosis of probable dengue and people requested the blood sample for confirmation of clinical diagnosis . The febrile patients ( n = 103/252 ) who did not consult a physician , 59 . 2% ( n = 61/103 ) mentioned previous contact with a dengue case in their neighborhood . Of all febrile patients that contact the dengue line , 78 . 82% ( n = 294 ) had 2 or more days with fever ( >38 °C ) . The age group that more frequently contact the FSD line was the 20–49 years group ( 41 . 5% ) , followed by the 5–9 years old ( 25% ) , >50 years ( 13 . 3% ) , 10–14 years ( 10 . 2% ) , 15–19 years ( 5 . 5% ) and 0–4 years old ( 4 . 5% ) . The average number of days between the beginning of the fever and the blood sample was 3 . 9±3 . 7 days , with significant variations between cities: Merida 4 . 4 days , Progreso 3 . 2 days and Ticul 2 . 8 days . Through the enhanced surveillance strategies ( absenteeism and toll-free number ) 244 serological tests for dengue ( IgM or NS1 ) were performed , 59% ( n = 144 ) from Merida , 17% ( n = 42 ) from Progreso and 24% ( n = 58 ) from Ticul . Only 8% of these samples were positive for dengue .
The results provided by the cohort of school children in three urban settings in the state of Yucatan , Mexico , described the high exposure to dengue infection in scholar groups that increased with age , without differences by sex but with significant differences between low , medium and high risk areas . The prevalence of antibodies in children 9 to 15 years above 60% in all areas except Progreso ( low risk ) confirmed the high transmission of dengue virus in this endemic state of the country . Prevalence found in this study were as expected and comparable to those from other endemic countries inside and outside the Americas , as reported in urban Nicaragua ( 2003 ) [24 , 25] or in urban settings in Central Brazil ( 2001 ) and northeast Brazil ( 2013 ) [26] with similar exposure and prevalence rates . Studies in India also showed increasing dengue infection with age and differences by region [27] were comparable to those reported in other endemic countries like Indonesia ( 1996 ) [28]; Sri Lanka ( 2014 ) [29] and Vietnam [30] . In the case of Yucatan , a state level prevalence of 72 . 5% was reported since 1985 [31]; seroprevalences in school children 8 to 14 years old ( 1987–1988 ) in urban ( 56 . 8% ) and rural Merida ( 63 . 7% ) [32] demonstrated the high exposure to dengue in the past , while another serological surveys done in Yucatan in 1996 and 2006 demonstrated seroprevalences of 22% and 20% in under five years old and 30% to 51% in 5 to 14 years old , respectively [33] . In other states of the country similar prevalence have been also reported: 35 . 7% in 5 to 9 years old and 52 . 2% for 10 to 14 years old ( 2011 ) [34] . Dengue transmission is highly heterogeneous and the burden varies by geographic region [35] , countries [36] , age groups affected and serotype [37] . Significant heterogeneity in transmission intensity has been identified within districts , sub-districts and even finer spatial scales like schools [38–41] . The differences between risk areas was expected since social , economic and environmental conditions have all proven to have certain influence in infection risks of populations [42 , 43] . These conditions could also influence transmission within the different schools in the selected areas . The low risk areas selected in our study comprise one urban area in the capital city and a town ( Progreso ) with historical low and occasional reports of dengue cases . Overall seroprevalence in these sites was significantly lower than the reported for medium and high risk areas . The higher prevalence of dengue infection in all age groups in Ticul ( medium risk ) demonstrated that conditions for transmission were wider than expected . The under report of cases by local health services as well as less demand for care from the sick population could be some distinctive traits within this community . Population movements between Ticul and high risk areas of Merida due to economic dynamics could also be a contributing factor . On the other hand , we did identify certain household risk conditions ( storage of water [44] , bathroom and kitchen outdoors ) in Ticul that could help explain this particular situation . Window screening has been suggested as an important feature to prevent Aedes from entering houses in Merida and potentially affect dengue transmission risk as well [45] . In this study , having ample coverage of screens in windows and doors was protective and urban programs should promote the inclusion of this preventive measure in houses in high risk areas [46 , 47] . Under reported cases is also a very common trait in school children in several settings [48 , 49] . The long history of dengue in Yucatan has created conditions where dengue is no longer identified as a health threat; fever is confused with other diseases and appears as a nonspecific febrile syndrome . The enhanced strategies introduced ( school absenteeism and toll-free number ) helped identify new cases within the school cohort although coverage and use of the telephone line needs to be promoted within the community to become a reliable and useful surveillance tool . Regular community home-to-home visits have proven to be an additional and useful strategy to support the traditional surveillance established by the health sector [50] . Circulation of DEN-1 , DEN2 and DEN-4 serotypes in the Yucatan region has not changed although the recent introduction of DEN-3 virus in 2016 could increase transmission due to low herd immunity towards this serotype . The introduction of chikungunya virus in 2015 and Zika virus in 2016 competed with all dengue serotypes and diminished the capacity to identify dengue cases by health providers and family members in the cohort as well . The enhance surveillance strategies implemented help improve detection of dengue cases under this circumstance . Limitations in our cohort are linked to the problems arising from the difficulties to engage family members and individuals in this kind of study . Losses to follow-up were higher in Merida and drop-out families were larger than those who stayed participating and it could generate under estimations of the risk of dengue in this city . The coincidence of chikungunya and Zika epidemics virus produced competing conditions for diagnosis and promoted intensive control interventions that eventually diminished our capacity to identify dengue cases or lower the transmission of dengue infection in the community . Nevertheless , the enhanced surveillance strategies allowed us to identify the three infections in our cohort . We did not perform entomological surveys to establish potential differences in vector densities by risk areas and these data could provide additional to support our findings . Screening of doors and windows may behave as a proxy of entomological risk since areas with this kind of protection are prone to have higher mosquito densities . With the potential introduction of innovative preventive or vector control interventions at sight , it is imperative that countries improve their surveillance system and produce baseline data that describe the epidemiological profile of the target population in order to improve the estimates of the direct and indirect effects of these individual or combined interventions . Our results confirmed the high exposure in these age groups and provide evidence that preventive and control interventions directed to children could decrease the burden of disease in high transmission areas [51] . | Dengue is a major public health problem in Latin America . Its transmission is highly heterogeneous , and its burden varies by geographic region , age group affected , serotype and other factors . While surveillance of dengue in the region has improved , several limitations remain , including under detection , misdiagnosis and the complexity of controlling a vector that has adapted to human dwellings in tropical and subtropical urban contexts . Prospective studies have become crucial to understand the transmission of dengue in urban environments and assess the impact of control strategies , such as the introduction of a dengue vaccine or additional vector control interventions . Our findings provide epidemiological data regarding the serological profile and risk factors for dengue infections in a cohort of children 0 to 15 years old in an endemic state in Mexico and confirmed the high exposure in these age groups . Likewise , enhanced and passive surveillance of cases gave us the opportunity to measure the behavior of dengue activity during chikungunya and Zika viruses’ arrival , which we believe will contribute to improve the design of surveillance and control strategies . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"dengue",
"virus",
"children",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"education",
"chikungunya",
"infection",
"pathogens",
"sociology",
"tropical",
"diseases",
"microbiology",
"social",
"sciences",
"viruses",
"age",
"grou... | 2018 | Dengue seroprevalence in a cohort of schoolchildren and their siblings in Yucatan, Mexico (2015-2016) |
Many repair and recombination proteins play essential roles in telomere function and chromosome stability , notwithstanding the role of telomeres in “hiding” chromosome ends from DNA repair and recombination . Among these are XPF and ERCC1 , which form a structure-specific endonuclease known for its essential role in nucleotide excision repair and is the subject of considerable interest in studies of recombination . In contrast to observations in mammalian cells , we observe no enhancement of chromosomal instability in Arabidopsis plants mutated for either XPF ( AtRAD1 ) or ERCC1 ( AtERCC1 ) orthologs , which develop normally and show wild-type telomere length . However , in the absence of telomerase , mutation of either of these two genes induces a significantly earlier onset of chromosomal instability . This early appearance of telomere instability is not due to a general acceleration of telomeric repeat loss , but is associated with the presence of dicentric chromosome bridges and cytologically visible extrachromosomal DNA fragments in mitotic anaphase . Such extrachromosomal fragments are not observed in later-generation single-telomerase mutant plants presenting similar frequencies of anaphase bridges . Extensive FISH analyses show that these DNAs are broken chromosomes and correspond to two specific chromosome arms . Analysis of the Arabidopsis genome sequence identified two extensive blocks of degenerate telomeric repeats , which lie at the bases of these two arms . Our data thus indicate a protective role of ERCC1/XPF against 3′ G-strand overhang invasion of interstitial telomeric repeats . The fact that the Atercc1 ( and Atrad1 ) mutants dramatically potentiate levels of chromosome instability in Attert mutants , and the absence of such events in the presence of telomerase , have important implications for models of the roles of recombination at telomeres and is a striking illustration of the impact of genome structure on the outcomes of equivalent recombination processes in different organisms .
Telomeres are the specific chromatin structures present at the ends of linear chromosomes [1] . They are known to play two main roles in the preservation of chromosomal integrity: avoiding terminal DNA sequence loss after replication and assuring that the chromosome ends are not recognized by the cellular machinery as DNA double-strand breaks [2]–[8] . In general , eukaryotic telomeres are composed of tandem repeats of a short sequence rich in G/C that terminates in a single strand 3′ overhang which can fold back and invade the duplex repeats to form the so called T-loop . A specific telomeric protein complex known as shelterin is implicated in the stabilization of the T-loop [9] , 10 . In mammalian cells this complex includes the specific telomeric-DNA-binding proteins TRF1 and TRF2 , which interact directly with duplex telomeric DNA , and POT1 which associates with the 3′ single stranded DNA . In most organisms telomeres are maintained by telomerase , a reverse transcriptase with a RNA subunit that serves as template for telomeric repeat synthesis . In the absence of telomerase , telomeres shorten with successive cell divisions , become non-functional and identified by the cell as damaged DNA , ultimately leading to genetic instability and cell death [11] , [12] . In recent years , many other proteins known for a more general role in cellular metabolism have been found to associate to telomeres , notably proteins involved in DNA repair and recombination . These include MRE11/RAD50/NBS1 , KU70/KU80 , DNAPKcs , BLM/WRN and ERCC1/XPF and have been found associated with telomeres and to play important roles in telomere protection and/or homeostasis ( reviews [5] , [7] , [13] , [14] ) . In the work presented here our interest has focussed particularly on the ERCC1/XPF heterodimer , which has been shown to associate to telomeres through interaction with TRF2 protein in mammalian cells [15] . ERCC1/XPF is a structure-specific endonuclease , initially identified for its essential role in nucleotide excision repair ( NER ) in budding yeast [16] . ERCC1 and XPF are highly conserved proteins and , in addition to yeast ( Rad1/Rad10 ) , orthologs have been identified in many organisms including Arabidopsis ( AtERCC1/AtRAD1 ) [17]–[22] , S . pombe ( Rad16/Swi10 ) [23] , [24] and Drosophila ( DmERCC1/MEI-9 ) [25] , [26] . The ERCC1/XPF endonuclease activity specifically recognises double- to single-strand transitions in DNA , incising the 5′–3′ single-strand just after the junction ( reviews by [27]–[30] ) . This DNA structure is a common element of homologous recombination intermediates and the 3′-ended G-strand overhang at telomeres is also a DNA structure of this type , although it is protected by the T-loop structure . In agreement with this , it has been shown that TRF2 is essential for T-loop stabilization , and its absence results in ERCC1/XPF -dependent , telomeric 3′ overhang loss [15] . Telomeres in most plant species are constituted of the repeat sequence TTTAGGG , initially identified in Arabidopsis thaliana [31] . Described plant telomeres vary in length from 2–9 Kb in Arabidopsis to 150 Kb in tobacco . The presence of G-overhangs has been detected in Arabidopsis and S . latifolia [32] and T-loops have been observed at telomeres of the garden pea , Pisum sativum [33] . Thus end-capping mechanisms seem to be conserved between mammals and plants . However , relatively little is known about plant telomeric proteins and in particular , the constituents of the plant shelterin complex have not been functionally identified [34] . Notwithstanding , a number of factors known for their roles in DNA repair such as the RAD50/MRE11 complex and the KU70/KU80 heterodimer has been found to play essential roles in protection of Arabidopsis chromosome ends [13] , [14] . Given the conserved functional roles of the mammalian ERCC1/XPF proteins and the plant orthologs AtERCC1/AtRAD1 in DNA repair and recombination , we present here an analysis of the roles of AtERCC1/AtRAD1 in telomere homeostasis and chromosomal stability in Arabidopsis plants . We demonstrate an essential role for the AtERCC1/AtRAD1 nuclease in the protection of shortened telomeres in Attert mutant plants . In striking contrast to XPF−/− and ERCC1−/− mammalian cells , Arabidopsis plants mutated for the AtERCC1 or AtRAD1 genes are viable and do not show any obvious defects in growth or development after more than 5 successive mutant generations . In the absence of telomerase , mutation of either AtERCC1 or AtRAD1 induces much earlier onset of developmental defects , correlated with increased genome instability . FISH analyses of mitotic anaphase figures shows that only 53% of the anaphase bridges in double mutant plants result from end-to-end chromosome fusions , compared to 91% in later generation Attert mutants with the same level of instability . Furthermore , 90% of the non end-to-end chromosome bridges are accompanied by large acentric DNA fragments in the double mutants . This simultaneous formation of a dicentric and an acentric chromosome is a consequence of recombination between telomeres and large interstitial blocks of degenerate telomeric sequences present on the right arms of chromosomes 1 and 4 . We conclude that the endonuclease AtERCC1/AtRAD1 protects short telomeres from “destructive” homologous recombination in Arabidopsis plants .
Absence of TRF2 protein in mammalian cells leads to telomere uncapping and chromosome fusions . Such fusions require the presence of the ERCC1/XPF nuclease , which by eliminating the single-stranded 3′ G-strand overhang , generates the non-homologous end-joining ( NHEJ ) substrate [15] . We decided to check whether the AtERCC1/AtRAD1 proteins are required for chromosome end-to-end fusions detected in the absence of telomerase in Arabidopsis plants [35] . To answer this question we generated double mutant Atercc1/Attert and Atrad1/Attert Arabidopsis lines and compared their phenotypes with those of single Atercc1 , Atrad1 and Attert mutant lines in successive generations . Homozygous Attert mutant plants were crossed to homozygous Atercc1 and to Atrad1 plants , to produce the doubly heterozygous F1 lines: Attert/AtTERT Atercc1/AtERCC1 and Attert/AtTERT Atrad1/AtRAD1 . Wild type , homozygous single Attert , Atercc1 , Atrad1 and double Attert/Atercc1 and Attert/Atrad1 F2 lines were selected and their growth and developmental phenotypes followed through successive generations of self-fertilisation . The original F2 lines are labelled Generation 1 ( G1 ) for the Attert mutant status , and successive generations labelled G2 , G3 , … . At any given generation , plants were identified as belonging to one of three arbitrary phenotypic classes: wild-type ( normal ) , semi-sterile ( reduced fertility ) or sterile ( this class includes plants arrested in vegetative growth and those unable to produce viable seeds ) ( Figure S1 ) . Single mutant Atercc1 and Atrad1 plants show wild-type phenotype and this is maintained over successive generations . Attert mutant plants show the expected progressive increase in both the proportion of plants presenting developmental defects and an increasing severity of these phenotypes over successive generations . The appearance and severity of these Attert phenotypes were however considerably advanced in the double Atercc1/Attert and Atrad1/Attert mutants . The results are presented in Figure 1A for the third , fourth and fifth ( G3 , G4 , G5 ) telomerase mutant generation plants ( see also Table S1 ) . G3 Atercc1/Attert seeds show 80% germination efficiency , compared to 100% in G3 Attert single mutant plants . More importantly , 17 . 9% of Atercc1/Attert plants were semi-sterile while no obvious defects were visible in the single Attert mutant plants . By generation five ( G5 ) , no normal Atercc1/Attert plants were observed from a total of 135 plants , while 68% of the G5 Attert mutant plants were phenotypically normal . Equivalent results were obtained for the Atrad1/Attert double mutant ( Table S1 ) . Thus , absence of the nuclease AtERCC1/AtRAD1 induces a substantial acceleration of the Attert-associated developmental phenotype in Arabidopsis plants . These observations raise the question of whether the accelerated developmental anomalies in Atercc1/Attert and Atrad1/Attert correlate with increased levels of cytogenetic damage in the double mutants . Successive generations of Attert mutant plants show progressive shortening of telomeres that eventually become uncapped and as a result end-to-end chromosomal fusions are generated . These fused , dicentric chromosomes can be detected as chromosome bridges at mitotic anaphase . We thus analyzed the frequencies of mitotic anaphase bridges in successive generations of the double and single mutant plants . For each mutant and generation , 200–300 mitotic anaphases were examined from pistil cells isolated from 3 individual plants ( Table S2 ) . As expected from their wild-type phenotype , no mitotic anaphases presenting bridges were detected in Atercc1 , nor in Atrad1 single mutant plants . Figure 1B presents the results for generations two to five ( G2–G5 ) of double Atercc1/Attert and single Attert mutant plants . No anaphase bridges were observed in cells from the three first generations of single Attert mutant plants . In contrast , Atercc1/Attert double mutant plants show 4–5% of anaphases with chromosome bridges in generation two and 15% in generation three . By generation four , 30% of the anaphases prepared from Atercc1/Attert plants show chromosome bridges , compared to only 2–5% in the single Attert mutant third generation plants . Moreover , a higher proportion of anaphases presenting 2 or 3 bridges were observed in double mutant plants compared to the Attert single mutant plants . Equivalent results were obtained in anaphase preparations from pistil cells from Atrad1/Attert double mutant plants in which chromosome bridges appear 3 generations earlier compared to the Attert single mutant lines derived from the same cross ( Table S2 ) . Thus the accelerated Attert phenotype observed in Attert plants lacking the AtERCC1/AtRAD1 nuclease is directly correlated with an earlier onset of genomic instability in these plants . These results strongly suggest a protective role of AtERCC1/AtRAD1 proteins at short telomeres generated in the absence of telomerase . This effect of the AtERCC1/AtRAD1 proteins contrasts with that observed in mammalian cells with uncapped telomeres due to lack of TRF2 , where the nuclease is essential for telomere fusion [15] . The simplest hypothesis to explain the accelerated appearance of genome instability is an increased rate of telomere erosion in the Atercc1/Attert double mutant plants . To test this hypothesis we carried out TRF analysis on DNA prepared from generations 2 to 5 of wild-type , Atercc1 , Atrad1 and Attert single mutants , and Atercc1/Attert and Atrad1/Attert double mutant plants ( Figure 2 ) . As expected from the absence of phenotype , telomeres of the Atercc1 and Atrad1 single mutant plants were maintained at the wild-type length through the four generations analyzed . A slight acceleration of telomere loss is observed in Atercc1/Attert and Atrad1/Attert double mutant plants , as compared to single Attert mutant plants . However , this cannot explain the appearance of fusions two generations earlier in double mutant plants . As shown in Figure 2 , telomeres are longer in G2 Atercc1/Attert plants than in G4 Attert mutant plants , although the former have a greater proportion of mitoses with bridges ( 3 . 4% ) than the latter ( 2 . 3% ) . Specific TRF analysis for the telomeres of the long arm of chromosome 2 and the short arm of chromosome 5 , confirmed that telomeres in Atercc1/Attert G2 are longer that in G4 Attert in contrast with the similar number of anaphases with bridges detected in these plants ( Figure 2 ) . Similar results were obtained in later generations and in Attert plants lacking the AtRAD1 protein ( Figure 2 and Table S2 ) . Thus , increased telomere erosion in the absence of the AtERCC1/AtRAD1 endonuclease Arabidopsis plants cannot explain the acceleration of telomere dysfunction in the Attert/Atercc1 and Attert/Atrad1 plants . The alternative hypothesis is that the AtERCC1/AtRAD1 proteins protect short telomeres against recombination . We thus carried out fluorescence in situ hybridisation ( FISH ) analyses of chromosome fusions using telomeric-repeat and subtelomeric probes ( a pool of BACs corresponding to the two ends of each of the five Arabidopsis chromosomes ) . The subtelomeric BAC FISH probes have been previously validated by Fibre-FISH [36] . Three categories of anaphase bridges could be detected in the FISH analyses , those corresponding to end-to-end chromosome fusions presenting subtelomeric signals with ( class I ) or without ( class II ) telomeric repeats , and bridges lacking both subtelomeric and telomeric signals ( class III ) ( Figure 2A ) . The proportions of the three classes of anaphase bridges were determined in Attert plants at G5 ( 10% of anaphases with bridges ) and G7 ( 25% anaphases with bridges ) . As expected , in correlation with the increased loss of telomere repeats in G7 plants , the proportion of bridges lacking telomeric signals is increased with respect to G5 plants ( 25% versus 6 , 5% ) . This increase was accompanied by a corresponding reduction in the proportion of bridges with telomeric repeats ( G5 84% , G7 66% ) . No changes were seen in the proportion of class III bridges lacking both subtelomeric and telomeric signals ( G5 9 , 5% , G7 9% ) ( Figure 3B ) . In contrast a substantial increase in the proportions of class II and III bridges was observed in Atercc1/Attert mitoses . Thus , G3 Atercc1/Attert cells that present a similar proportion of anaphases with bridges to G5 Attert cells ( 10–15% ) , show 49% of class II ( versus 6 , 5% in Attert cells ) and 18% of class III ( versus 9% in Attert cells ) ( Figure 3B ) . Strikingly , the proportion of class III bridges increases up to 47% in Atercc1/Attert G5 cells , versus 9% in G7 Attert cells which show the same proportion of anaphases with bridges ( 25% ) . Thus , by G5 almost half of the anaphase bridges in Atercc1/Attert cells do not result from chromosome end-to-end fusions . Similar results were obtained for Atrad1/Attert cells ( data not shown ) . These observations suggest strongly that AtERCC1/AtRAD1 endonuclease protects short telomeres against other types of recombination than the fusion of uncapped chromosomes ends . Careful observation of the FISH images of mitotic Atercc1/Attert cells revealed extrachromosomal DNA masses in 90% of the anaphases presenting class III bridges ( Figure 4 ) . In all cases this extrachromosomal DNA hybridized both with subtelomeric and telomeric repeat FISH probes . To better understand the nature of these extrachromosomal DNAs , we realized FISH analyses using a pool of 10 BAC probes situated in the middle of each arm of the five Arabidopsis chromosomes . All the analyzed anaphase figures containing extrachromosomal DNA produced a positive signal . Identical results were obtained with a mix of centromere-proximal BAC probes specific for each chromosome arm . Equivalent results were obtained for Atrad1/Attert double mutant plants ( data not shown ) . Class III bridges in Atercc1/Attert and Atrad1/Attert are thus associated with the generation of acentric DNA corresponding to a chromosome arm in at least 90% of cases . In order to explain these data , we considered the hypothesis that in the absence of AtERCC1/AtRAD1 , short telomeres unable to form a T-loop could invade internal telomere-related sequences , in a similar manner to the events proposed to generate telomeric double minute chromosomes in ERCC1-deficient mouse cells [15] . With only 5 chromosome pairs and a fully sequenced genome , Arabidopsis is a particularly good model for the dissection of such events . We thus used the sequence viewer of the Arabidopsis Information Resource ( TAIR ) ( http://www . arabidopsis . org/servlets/sv ) and the “fuzznuc” program of the EMBOSS suite [37] to map and characterise interstitial telomeric repeats in the Arabidopsis genome , presented in schematic form in Figure 5 . The bioinformatics search of sequences revealed the presence of 4 contiguous perfect TTTAGGG repeats on chromosomes 1R and 2L and three repeats at two loci on chromosome 5R . Searching for the CCCTAAA sequence identified three loci with 8 , 30 and 5 contiguous perfect repeats on chromosome arms 1R , 3L and 4R respectively . Five loci of three contiguous repeats are found on chromosomes 1L , 3L , 3R and 4R . Of particular interest are two extensive regions of degenerate telomeric repeats identified on chromosome 1R ( 349 Kb ) and 4R ( 67 Kb ) . 5 . 9% of the chromosome 1R region DNA consists of perfect C-strand telomere repeats ( CCCTAAA ) n , a figure which rises to 17 . 2% if repeats with 1 mismatch are included . G-strand repeats ( TTTAGGG ) n are considerably less represented in this block , with 0 . 05% and 1 . 3% respectively . Similarly the chromosome 4R region has 8 . 8% ( perfect ) and 25 . 18% ( including 1 mismatch ) C-strand repeats and 0 ( perfect ) and 0 . 8% ( 1 mismatch ) G-strand repeats . The longest perfect tandem C-strand repeats in these regions are an 8-mer in the 1R region and a 5-mer in the 4R region . On the G-strand , the longest perfect tandem repeat in these regions is a 3-mer in the 1R region , with none in the 4R region . Should a telomeric 3′-ended G-overhang recombine and crossover with interstitial telomeric repeat sequences , the consequences would depend upon the orientation of these sequences relative to the centromere ( s ) and whether or not they are on the same chromatid or chromosome arm . In all , there are eight possible invasion configurations ( Figure 6A ) . Holliday junction resolution of only two of these would generate the observed class III dicentric chromosome ( anaphase bridge lacking subtelomeric and telomeric repeats ) and an acentric chromosome , as illustrated in Figure 6B and 6C . Figure 6B shows invasion by another chromosome ( or chromatid ) of ( CCCTAAA ) n sequences , such as those present in the extensive blocks on the right arms of chromosomes 1R and 4R . Figure 6C shows invasion by another chromosome ( or chromatid ) of the ( TTTAGGG ) n sequence , such as that present on the left arm of chromosome 2 . In both cases resolution of the resulting Holliday junction can give rise to a dicentric and an acentric chromosome . According to this model , the acentric DNA observed in Atercc1/Attert Arabidopsis cells should correspond to chromosome arms 1R , 4R and possibly , 2L . To test this prediction , we carried out FISH analyses using BAC probes corresponding to all 10 individual Arabidopsis chromosome arms . The results in Figure 7A show that only probes for the right arms of chromosomes 1 and 4 hybridized with the extrachromosomal DNA . We confirmed these results with FISH using only the probes to Chr . 1R ( red ) and Chr . 4R ( green ) – the extrachromosomal DNA in 35 out of the 36 anaphases examined hybridised to one of these two probes ( Figure 7B , C ) . The acentric fragments thus correspond to the chromosome arms distal to the two extensive blocks of interstitial telomeric DNA and these data thus strongly support the origin of the observed dicentric+acentric chromosomes in Atercc1/Attert plants through homologous recombination of telomeric and interstitial telomere-repeat DNA sequences .
We present here an analysis of the roles of the structure-specific ERCC1/XPF ( AtERCC1/AtRAD1 ) endonuclease in telomere homeostasis in the plant Arabidopsis thaliana . Double Atercc1/Attert or Atrad1/Attert mutant Arabidopsis plants show considerably more severe growth and developmental phenotypes than single Attert mutant plants . This aggravation of the telomerase mutant phenotype is directly correlated with an earlier onset of chromosome instability , as detected by the appearance of mitotic dicentric anaphase bridges . Analysis of the structure of these dicentric chromosomes shows that , in contrast to Attert plants where 90% of dicentrics result from end-to-end fusion , 50% of dicentrics in Atercc1/Attert cells result from recombination of telomeres with two extensive regions of interstitial telomere-related DNA in the Arabidopsis genome . In mammalian cells the ERCC1/XPF heterodimer is associated to telomeres through interaction with TRF2 [15] . The rapid ageing phenotype of ERCC1−/− and XPF−/− mice has been attributed to roles of these proteins in DNA repair mechanisms other than NER , such as the repair of DNA interstrand cross-links ( 44 , 45 ) and DSB ( 46 , 47 ) . Cytogenetic analyses of ERCC1−/− mouse embryonic fibroblasts ( MEFs ) show neither defects in telomeres nor in telomeric G-strand homeostasis and no end-to-end chromosomes fusions were detected in these cells . However , FISH analysis on metaphase spreads of ERCC1−/− MEF cells showed greatly elevated numbers of telomere-containing double-minute chromosomes ( TDMs ) , compared to wild-type and XPC−/− controls [15] . This generation of double minute chromosomes presumably contributes to the severe postnatal growth defects and death at 3 weeks of mice mutated for either protein [38]–[42] . Although the roles of the ERCC1/XPF nuclease in NER and double strand break repair are conserved in Arabidopsis [17] , [18] , [21] , Arabidopsis Atercc1 and Atrad1 mutants grow and develop normally and show no detectable chromosomal instability , with neither anaphase bridging nor alterations in bulk telomere length detected in these plants after more than five mutant generations ( this work ) . The situation is however strikingly different in plants also lacking telomerase , in which absence of either the AtERCC1 or AtRAD1 proteins dramatically advances the appearance of developmental defects and chromosomal instability . As with many other organisms including mammals , absence of telomerase leads to progressive shortening of telomeric repeat arrays , destabilising the T-loop structure at telomeres and resulting in their recognition by the cellular recombination machinery . Recombination of these shortened , non-functional telomeres principally results in end-to-end chromosomal fusions and dicentric chromosomes ( reviewed by [10] ) . However , overhanging G-strand telomeric DNA from non-functional telomeres could also invade and recombine with interstitial telomere-like sequences . Such recombination between telomeres and interstitial sequences would have differing consequences , depending on the location and orientation of these interstitial sequences with respect to the centromere . Studies in cultured human cells suggest that the ERCC1/XPF endonuclease would play two roles in the avoidance of such events: removal of G-strand overhangs at decapped telomeres would reduce the propensity of these to invade cognate interstitial sequences and should such invasion occur , ERCC1/XPF cleavage of the intermediate structure would pre-empt resolution by the recombination machinery [15] . In this work we report strong developmental phenotypes in Atercc1/Attert ( and Attert/Atrad1 ) Arabidopsis plants , associated with the frequent occurrence of mitoses with dicentric and acentric chromosomes . Southern analysis shows little or no acceleration of bulk telomere shortening in the double Atercc1/Attert and Attert/Atrad1 mutants , compared to single Attert plants . Thus , although a minor contribution of the absence of AtERCC1/AtRAD1 to telomere shortening in Attert mutants cannot be ruled out , this cannot explain the dramatic acceleration of the developmental and chromosomal instability phenotypes observed in the double mutants . This conclusion is reinforced by the striking decrease in the relative proportion of end-to-end ( Class I ) chromosomal fusions in double Atercc1/Attert and Attert/Atrad1 mutants compared to single Attert mutants . This is more than compensated by relative increases in proportions of mitoses with dicentric bridges lacking telomeric ( Class II ) or both telomeric and sub-telomeric ( Class III ) DNA in double mutants . Furthermore , acentric chromosome arms are observed in 90% of mitoses with Class III dicentric bridges in the double mutants . Absence of AtERCC1/AtRAD1 thus both strongly increases the numbers and affects the nature of chromosomal fusions in Attert mutants . In order to determine whether these dicentric+acentric figures could result from recombination between telomeres and interstitial telomeric-sequences , we analysed the numbers and positions of internal telomere-related sequences in the fully sequenced Arabidopsis genome . This bioinformatics analysis shows the presence of 2 extensive blocks of degenerate telomere sequence on the right arms of Arabidopsis chromosomes 1 and 4 . Invasion of these interstitial telomere-related sequences by the G-strand of a decapped telomere would create a recombination intermediate , processing of which by the homologous recombination machinery could generate the observed dicentric+acentric mitotic figures ( Figure 8: B–>E–>G ) . Our analysis predicts that the acentric chromosomes should correspond to the right arms of either chromosome 1 or 4 , a prediction confirmed in 35/36 of such acentrics examined . The processes leading to the events which we describe in Arabidopsis plants ( dicentric chromosome+acentric arm; Figure 8G ) thus appear equivalent to those resulting in the TDMs in human cells described by Zhu et al [15] . The striking difference between our data and that in animal cells is the normal growth and development and the absence of karyotypic abnormalities in ( AtTERT+ ) Atercc1 or Atrad1 mutant plants . The karyotypic instability of Atrad1 and Atercc1 mutants thus depends upon the absence of telomerase . This implies that interstitial telomere invasions do not occur , or are very rare in wild-type plants , in contrast to the observations in cultured human cells , where such invasion events are presumably very frequent ( TDMs observed in 44–86% of mitoses in different ERCC1−/− cell lines ) . It is striking to note the different outcomes of equivalent recombination processes in the different organisms . Homologous recombination of telomeres with interstitial sequences in direct orientation on the same chromatid arm is proposed to generate massive chromosome breakage and circular acentrics ( TDMs ) in mouse cell culture , while recombination of telomeres with interstitial sequences in inverted orientation on another chromatid arm leads to breakage and fusion of two specific chromosome arms in Arabidopsis . These differences thus appear to primarily depend upon the structure of the genome/karyotype and the locations of the interacting sequences . This striking illustration of the different outcomes and impacts of recombination processes in the different genomic contexts is underlined by recent data from the fission yeast , S . pombe [43] . Absence of telomerase ( Trt1 ) in S . pombe leads to telomere shortening and cell death . However , rare “survivor” cells escape and grow normally due to circularisation of their chromosomes . A recent report shows that the absence of Rad16 has no effect on the rate of telomere shortening in trt1 cells , but strongly reduced the occurrence of survivors ( Rad16 is the S . pombe XPF ortholog ) . In a series of elegant experiments , these authors further show that the chromosome circularisation leading to survival of trt1 cells occurs through Rad16-dependent , single-strand annealing ( SSA ) recombination between homology regions present as inverted repetitions between 7 and 13 Kb from the telomeres of chromosomes I and II [43] . As in Arabidopsis and animal cells , absence of S . pombe Rad16 protein thus profoundly affects the recombination of de-protected telomeres , at least in this selected subset of “survivor” events . In Arabidopsis Atercc1/Attert plants , end-to-end chromosome fusions represent 53% of ( total ) anaphase bridges , which thus cannot have been generated through the ERCC1/XPF -dependent SSA recombination pathway . We are currently initiating work to elucidate the roles of the different homologous and non-homologous recombination pathways in the generation of these fusions .
Arabidopsis thaliana plants were grown in soil in the greenhouse under standard conditions . The Attert [44] , Atercc1 [18] and Atrad1 [45] Arabidopsis mutants have been described previously . The two double Atrad1/Attert and Atercc1/Attert mutants were produced by crossing Atrad1 and Atercc1 homozygotes with an Attert homozygote ( 3rd mutant generation ) , using standard techniques . PCR genotyping was carried out as described for Attert [44] , Atrad1 [19] and Atercc1 [18] . TRF analysis of telomere length in Mbo1-digested genomic DNA was as previously described for the telomeric and subtelomeric chromosome 2 probes [46] and for subtelomeric 5 probe [47] . Whole flower buds were collected and fixed , pistils were digested and were squashed on slides [48] . Slides were mounted using Vectashield ( Vector Laboratories ) mounting medium with 1 . 5 µg/ml DAPI ( 4′ , 6-Diamidino-2-Phenylindole ) and observed by fluorescence microscopy , using a Zeiss Imager . Z1 microscope . Images were further processed and enhanced using Adobe Photoshop software . BACs from subtelomeric regions of Arabidopsis chromosomes ( F6F3 , F23A5 , F15B18 , F17A22 , F4P13 , T20O10 , F6N15 , T19P19 , F7J8 , K9I9 ) , the middle of chromosome arms ( F12K11 , F20D21 , T8K22 , F12C20 , K1G2 , F16L2 , T5K18 , T1A4 , MIJC20 ) [49] and centromere-proximal regions ( F12K21 , F2J6 , T25N22 , T10F5 , T4A2 , T5C2 , T32N4 , T32A17 , T8M17 , F5H8 ) [49] were labelled with biotin-16-dUTP or digoxigenin-11-dUTP ( Roche ) using the BioPrime DNA labelling system ( Invitrogen ) and telomeric probe was labelled by PCR [ ( 95°C 1′ , 55°C 40″ , 72°C 2′ ) ×5 ( 94°C 1′ , 60°C 40″ , 72°C 2′ ) ×25] with digoxigenin-11-dUTP using specific telomere primers 5′ ( TTTAGGG ) 63′ . FISH experiments were performed according to [50] , as previously described [36] . For the detection of biotin-labelled probes , avidin∶Texas Red ( 1∶500 , Vector Laboratories ) followed by goat anti-avidin∶biotin ( 1∶100 , Vector Laboratories ) and avidin-Texas Red ( 1∶500 ) were used . Mouse anti-digoxygenin ( 1∶125 , Roche ) followed by rabbit anti-mouse∶fluorescein isothiocyanate ( FITC ) ( 1∶500 , Sigma ) and goat anti-rabbit∶Alexa 488 ( 1∶100 , Molecular Probes ) were used for the detection of digoxygenin-labelled probe . For multiple hybridisations of the same slide , FISH was carried out according to Mokros et al [51] , using BACs labelled either with Cy5-dUTP or Cy3-dUTP ( Amersham ) by standard nick translation reactions ( Roche ) . | Telomeres are the specialised nucleoprotein structures evolved to avoid progressive replicative shortening and recombinational instability of the ends of linear chromosomes . Notwithstanding this role of telomeres in “hiding” chromosome ends from DNA repair and recombination , many repair and recombination proteins play essential roles in telomere function and chromosome stability . Among these are XPF and ERCC1 , which form a structure-specific endonuclease known for its essential role in nucleotide excision repair and that is the subject of considerable interest in studies of recombination . In this study , we analyse the roles of the XPF/ERCC1 in telomere function and chromosome stability in the plant Arabidopsis thaliana , which , with its remarkable tolerance to genomic instability and sequenced genome , is an excellent higher eukaryotic model for these studies . Surprisingly , and in striking contrast to observations in mammalian cells , we observe no enhancement of chromosomal instability in Arabidopsis plants lacking either of these two proteins , which develop normally and show wild-type telomere length . However , Atercc1 ( and Atrad1 ) mutants profoundly affect the recombination of de-protected telomeres , dramatically potentiating chromosome instability . These results provide a striking illustration of the different outcomes and genomic impacts of the same recombination processes in different organisms . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics",
"genetics",
"and",
"genomics/plant",
"genetics",
"and",
"gene",
"expression",
"genetics",
"and",
"genomics/chromosome",
"biology"
] | 2009 | ERCC1/XPF Protects Short Telomeres from Homologous Recombination in Arabidopsis thaliana |
Vibrio parahaemolyticus is one of the human pathogenic vibrios . During the infection of mammalian cells , this pathogen exhibits cytotoxicity that is dependent on its type III secretion system ( T3SS1 ) . VepA , an effector protein secreted via the T3SS1 , plays a major role in the T3SS1-dependent cytotoxicity of V . parahaemolyticus . However , the mechanism by which VepA is involved in T3SS1-dependent cytotoxicity is unknown . Here , we found that protein transfection of VepA into HeLa cells resulted in cell death , indicating that VepA alone is cytotoxic . The ectopic expression of VepA in yeast Saccharomyces cerevisiae interferes with yeast growth , indicating that VepA is also toxic in yeast . A yeast genome-wide screen identified the yeast gene VMA3 as essential for the growth inhibition of yeast by VepA . Although VMA3 encodes subunit c of the vacuolar H+-ATPase ( V-ATPase ) , the toxicity of VepA was independent of the function of V-ATPases . In HeLa cells , knockdown of V-ATPase subunit c decreased VepA-mediated cytotoxicity . We also demonstrated that VepA interacted with V-ATPase subunit c , whereas a carboxyl-terminally truncated mutant of VepA ( VepAΔC ) , which does not show toxicity , did not . During infection , lysosomal contents leaked into the cytosol , revealing that lysosomal membrane permeabilization occurred prior to cell lysis . In a cell-free system , VepA was sufficient to induce the release of cathepsin D from isolated lysosomes . Therefore , our data suggest that the bacterial effector VepA targets subunit c of V-ATPase and induces the rupture of host cell lysosomes and subsequent cell death .
Vibrio parahaemolyticus , a Gram-negative marine bacterium , is a major food-borne pathogen that causes acute human gastroenteritis associated with the consumption of seafood . This pathogen also causes wound infections and septicemia in humans [1]–[3] . V . parahaemolyticus possesses virulence factors such as thermostable direct hemolysin ( TDH ) and two separate type III secretion systems ( T3SSs ) , namely , T3SS1 and T3SS2 [4] , [5] . T3SSs are protein export systems that enable bacteria to secrete and translocate proteins known as effectors into the cytoplasm of host cells . Translocated effectors modify host cell function and allow pathogens to promote infection and cause disease [6] , [7] . V . parahaemolyticus T3SS1 is involved in the cytotoxicity to various mammalian cells , whereas T3SS2 is related to the enterotoxicity of this organism [8]–[10] . T3SS1-induced cell death occurs rapidly , within several hours after the inoculation of V . parahaemolyticus; is independent of caspases; and is associated with autophagy [11] , [12] . The transcription of the T3SS1 genes of V . parahaemolyticus is regulated by a dual regulatory system consisting of the ExsACDE regulatory cascade and H-NS [13] . To date , VepA ( VP1680/VopQ ) , VopS ( VP1686 ) and VPA450 have been identified as T3SS1 effectors [14]–[18] . VopS functions as an AMPylator and contributes to cell rounding [17] . VPA450 acts as an inositol phosphatase and induces plasma membrane blebbing [18] . It has previously been shown that the deletion of vopS or vpa450 has little effect on T3SS1-dependent acute cytotoxicity , but a mutant strain of V . parahaemolyticus in which vepA was deleted ( ΔvepA ) lost its cytotoxicity , suggesting that VepA plays a major role in T3SS1-dependent cytotoxicity [14]–[16] . VepA is a 50-kDa protein and consists of 492 amino acids . Although VepA has no homology to known proteins , the amino-terminal 100 amino acids of VepA are required for secretion by T3SS1 [15] . VepA has also been reported to be involved in the activation of autophagy and mitogen-activated protein kinases ( MAPKs ) [16] , [19] . However , the mechanism by which VepA is involved in acute cytotoxicity in host cells is unclear . To understand the mechanism underlying the T3SS1-dependent cytotoxicity of V . parahaemolyticus , elucidation of the function of VepA within host cells is important because VepA appears to play a critical role in cytotoxicity . In this study , we showed that VepA itself is a cytotoxic effector , and we screened for host factors essential for the cytotoxicity of VepA using yeast genome-wide analysis to elucidate the function of VepA .
To understand the function of VepA , we first examined the expression of green fluorescent protein ( GFP ) in 293T cells transfected with pEGFP-VepA . However , GFP-VepA expression was not detected ( Figure 1A ) , raising the possibility that VepA itself is quite toxic in cells . To evaluate whether VepA itself is cytotoxic , we introduced purified VepA into HeLa cells by protein transfection using the hemagglutinating virus of Japan ( HVJ ) envelope vector . The delivery of VepA into cells caused a significant decrease in cell viability , in contrast to the delivery of glutathione-S-transferase ( GST ) , bovine serum albumin ( BSA ) or HVJ alone ( Figure 1B ) . VepA did not affect cell viability in the absence of the HVJ vector . These results suggest that VepA itself is cytotoxic and effective only inside cells , not outside cells . By contrast , the delivery of a truncated VepA protein lacking the C-terminal 92 amino acids ( 1–400 , VepAΔC ) into HeLa cells did not affect the viability of the cells ( Figure 1B ) . Although complementation of the wild-type vepA gene in the ΔvepA strain ( POR-3ΔvepA/vepA ) rescued the infection-mediated cytotoxicity , complementation with vepAΔC ( POR-3ΔvepA/vepAΔC ) did not ( Figure 1C ) . We also confirmed that VepAΔC is secreted from POR-3ΔvepA/vepAΔC ( Figure 1D ) , excluding the possibility that VepAΔC is not expressed or not secreted in V . parahaemolyticus . GFP-VepAΔC was successfully expressed in 293T cells , in contrast to GFP-VepA ( Figure 1A ) , suggesting that VepAΔC is stable in cells . Taken together , these results indicate that VepAΔC loses the cytotoxicity . To investigate the mechanism underlying VepA-dependent cytotoxicity , we tested whether autophagy and the MAPK signaling pathway were required for the cytotoxicity induced by V . parahaemolyticus . Inhibiting autophagy or MAPK with short interfering RNAs ( siRNAs ) or inhibitors did not affect the cytotoxicity ( Figure S1 ) , indicating that the contributions of autophagy and the MAPK signaling pathway to cytotoxicity are negligible . The cell-permeable pan-caspase inhibitor carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone ( Z-VAD-FMK ) also did not affect the cytotoxicity , a result that is consistent with those of a previous study which the V . parahaemolyticus strain NY-4 was used [12] . Saccharomyces cerevisiae has been widely used as a model system to study eukaryotic cells . An increasing number of reports have shown that the expression of bacterial effectors inhibits yeast growth , and this inhibition is implicated in the activity of effectors that affect cellular processes conserved among eukaryotic cells [20] . To determine whether VepA can inhibit yeast growth , we transformed S . cerevisiae BY4730 with the p426 expression plasmid encoding VepA and expressed VepA under the control of the GAL1 promoter . The ectopic expression of VepA inhibited the growth of yeast ( Figure 2A ) , indicating that VepA is also toxic in yeast . By contrast , VepAΔC , which lacks cytotoxicity ( Figure 1A , 1B and 1C ) , was less toxic in yeast , suggesting a correlation between the cytotoxic effects of VepA in HeLa cells and its toxicity in yeast . Next , we used a yeast knockout ( YKO ) strain library [21] to screen for host genes that are essential for the toxicity of VepA ( Figure 2B ) . To express VepA in non-essential gene mutant yeast strains and screen for clones that are able to grow in the presence of VepA , we transformed 56 pools of YKO strains ( one pool typically includes 95 strains ) with the p426 expression plasmid encoding VepA and plated the yeast onto SC plates lacking uracil and containing galactose ( SC-Ura+Gal ) . To examine the plating efficiency , the transformants were also plated onto SC plates lacking uracil and containing glucose ( SC-Ura+Glc ) , yielding at least 1 , 000 colonies ( giving >10×coverage ) . Using this genome-wide screen , we found that the Δvma3 strain was insensitive to the toxicity of VepA ( Figure 2C ) . No MAPK- or autophagy-related gene mutants were identified in this screen , a result that is consistent with the cytotoxicity analysis presented in Figure S1 . The expression of VepA in the Δvma3 strain was confirmed under inducing conditions ( Figure 2D ) . The ectopic expression of VopT and VopP , which have been reported to inhibit yeast growth [9] ( Figure 2A ) , was toxic to the Δvma3 strain ( Figure 2C ) , thus excluding the possibility that the Δvma3 strain was non-specifically insensitive to bacterial effectors . Complementation of the VMA3 gene in the Δvma3 strain restored the susceptibility to the toxicity of VepA ( Figure S2A ) . VMA3 encodes subunit c of V-ATPase . V-ATPases , which are highly conserved in eukaryotic cells , are composed of two domains: a peripheral , catalytic V1 domain and an integral V0 domain that serves as the basal body . V-ATPases serve as proton pumps to acidify intracellular compartments [22] . In yeast , the V1 and V0 domains contain eight ( A , B , C , D , E , F , G and H ) and six ( a , c , c′ , c″ , d and e ) subunits , respectively . Because the deletion of any subunit causes a functional deficiency in V-ATPases [23] , we determined whether the function of the V-ATPase is involved in the VepA-mediated growth defects in yeast . VepA was expressed in the yeast strains containing mutations of each component of the V-ATPase , and their growth was observed . With the exception of Δvma3 , none of the yeast strains with mutations in the V-ATPase subunits were able to grow in the presence of VepA ( Figure 2E ) , thus indicating that the function of the V-ATPase is dispensable for the toxicity of VepA in yeast . Next , we investigated the involvement of ATP6V0C , the human ortholog of yeast Vma3p , in VepA-mediated cytotoxicity . HeLa cells were transfected with siRNAs to knockdown ATP6V0C or ATP6V1A , subunit A of human V-ATPase . The transfected cells were infected with V . parahaemolyticus strain POR-3 [14] , and the level of cytotoxicity was evaluated . The knockdown of ATP6V0C reduced infection-mediated cytotoxicity , but the knockdown of ATP6V1A did not ( Figure 3A ) . The knockdown of expression of each protein was confirmed by immunoblotting ( Figure 3B ) . We also determined whether the knockdown of ATP6V0C affected the translocation of VepA in cells . HeLa cells treated with siRNAs were infected with POR-3 , ΔT3SS1 or ΔvepA strain for 3 h . The cell lysates were then analyzed by immunoblotting . The amount of VepA that was associated with the cells was not different between the control siRNA-treated cells and the ATP6V0C siRNA-treated cells ( Figure 3C ) , indicating that the knockdown of ATP6V0C did not affect the translocation of VepA into cells . Moreover , pre-treatment with the pharmacological V-ATPase inhibitors , bafilomycin A and concanamycin A [22] at 100 nM , a concentration that is sufficient to prevent vesicular acidification as validated by LysoTracker staining ( data not shown ) , did not prevent infection-mediated cytotoxicity ( Figure 3D ) . These results indicate that ATP6V0C is involved in VepA-mediated cytotoxicity and that the function of the V-ATPase is not required for cytotoxicity , which is consistent with the yeast study described above ( Figure 2E ) . Next , we characterized the localization of VepA in V . parahaemolyticus-infected HeLa cells by biochemical subcellular fractionation . HeLa cells were infected with V . parahaemolyticus strains for 3 h to avoid the severe cytotoxicity that occurred at 4 h and fractionated into cytosolic , membrane/organelle , nuclear and cytoskeleton fractions . Each of the fractions was then analyzed by immunoblotting with an anti-VepA antibody . VepA was localized in the membrane/organelle fractions of cells infected with POR-3 or POR-3ΔvepA/vepA ( Figure S3A ) . Furthermore , we fractionated the infected cells by ultracentrifugation ( Figure S3B ) . VepA was mainly detected in the upper fractions , predominantly fraction 1 , of cells infected with POR-3 or POR-3ΔvepA/vepA . This distribution of VepA is similar to that of V-ATPase ( ATP6V1A and ATP6V0D1 ) , which suggests that VepA may be associated with V-ATPases in infected cells . We next investigated whether VepA interacts with ATP6V0C . Biotinylated VepA ( b-VepA ) ( Figure S4 ) was adsorbed onto streptavidin beads and incubated with lysates of 293T cells expressing ATP6V0C-Flag . Immunoblot analysis showed that ATP6V0C-Flag co-precipitated with b-VepA-immobilized beads ( Figure 4A ) , indicating that VepA interacts with ATP6V0C . To further search for endogenous proteins that associate with VepA , proteins bound to VepA were pulled down from lysates of RAW264 . 7 cells , which are susceptible to T3SS1-induced cytotoxicity , and thus , have a potential for high expression of target molecules for VepA [10] , and visualized by silver staining of SDS-PAGE gels ( Figure 4B ) . We found a protein with a molecular weight of approximately 16 kDa that specifically associated with VepA . This protein band was excised , analyzed by liquid chromatography-tandem mass spectrometry ( LC-MS/MS ) and identified as ATP6V0C , as expected . By contrast , VepAΔC did not interact with Flag-tagged or endogenous ATP6V0C ( Figure 4A and 4B ) . In addition , in cells infected with POR-3ΔvepA/vepAΔC , the subcellular distribution of VepAΔC was different from that of VepA , and the protein was not enriched in fraction 1 , which contains lysosomes ( Figure S3B ) . ATP6V0C is an integral membrane protein that consists of 155 amino acids and is predicted to possess four transmembrane domains and two cytoplasmic loops [24] ( Figure S5A ) . VepA is injected into the cytoplasm of host cells by the T3SS and therefore might target the cytoplasmic loops of ATP6V0C . To test this possibility , 293T cells expressing the cytoplasmic loops ( Vc1: ATP6V0C 31–80 or Vc2: ATP6V0C 111–155 ) were infected with V . parahaemolyticus . Interestingly , the expression of Vc1 significantly reduced infection-mediated cytotoxicity ( Figure S5B ) . In addition , Vc1 interacted with VepA in pull-down assays ( Figure S5C ) . Taken together , our data indicate that ATP6V0C is a primary cellular target for VepA . V-ATPases are expressed on the membranes of various intracellular organelles , where they transport H+ across the membrane to generate and maintain acidic compartments . The lysosome is a major acidic compartment that contains various hydrolytic enzymes and functions as a digestive apparatus . The leakage of degradative contents from lysosomes into the cytosol is known to induce cell death , and the type of cell death that occurs is thought to be dependent on the extent of lysosomal damage , i . e . , partial and moderate ruptures cause apoptosis , whereas more severe damage leads to necrosis [25] , [26] . To investigate the integrity of lysosomes in infected cells , we used acridine orange ( AO ) , which is a fluorochrome stain used for vital staining of lysosomes that exhibits red fluorescence when concentrated in lysosomes and green fluorescence when diffused in the cytosol [27] . AO-loaded HeLa cells were infected with V . parahaemolyticus . Enhanced green fluorescence was observed in cells infected with POR-3 or ΔvepA/vepA but not in cells infected with POR-3ΔvepA or ΔT3SS1 , both of which are deficient in VepA ( Figure 5A–I ) . Thus , the relocation of AO to the cytosol was VepA dependent . To examine the extent of lysosomal leakage , we determined whether cathepsin D ( CatD ) , a lysosomal aspartyl protease , translocates from lysosomes to the cytosol . CatD was detected in the cytosolic fractions of cells infected with POR-3 or POR-3ΔvepA/vepA but not in cells infected with VepA-deficient strains or POR-3ΔvepA/vepAΔC , indicating that CatD was also released into the cytosol during infection in a VepA-dependent manner ( Figure 5K ) . Furthermore , the knockdown of ATP6V0C partially reduced the release of CatD into the cytosol during infection ( Figure 5L ) . We also treated AO-loaded HeLa cells with TDH , a Vibrio exotoxin . Although TDH does not contribute to the cytotoxicity of V . parahaemolyticus during infection , TDH attacks plasma membranes and functions as a pore-forming toxin that leads to cell death when HeLa cells are exposed to a high concentration of purified TDH [10] , [28] . In contrast to infection with V . parahaemolyticus , treatment with TDH did not induce the relocation of AO to the cytosol ( Figure S6A–D ) . Taken together , these results suggest that VepA-induced lysosomal rupture precedes plasma membrane disruption . In addition , pre-treatment of cells with U18666A or deferoxamine ( DFO ) , both of which increase lysosomal membrane stability [29] , [30] , reduced infection-mediated cytotoxicity ( Figure S6E and S6F ) . We next investigated the involvement of VepA in infection-mediated lysosomal rupture . Because VepA interacts with V-ATPase subunit c , which is highly expressed on lysosomal membranes , it is possible that VepA directly affects the integrity of lysosomes . We therefore examined whether VepA could directly rupture lysosomes in a cell-free system . Isolated lysosomes were incubated with VepA or VepAΔC . After precipitating the lysosomes , the release of CatD from the lysosomes into the supernatant was examined . For lysosomes treated with VepA , detectable levels of CatD were released into the supernatant . By contrast , the release of CatD was not observed in the supernatants from lysosomes treated with VepAΔC ( Figure 5M ) . Therefore , these results suggest that the cytotoxic T3SS effector VepA alone is able to induce lysosomal rupture and the leakage of contents from lysosomes . Finally , we examined the effect of knockdown of ATP6V0C on VepA-induced lysosomal rupture in the cell-free system . Lysosomes that were isolated from control siRNA- or ATP6V0C siRNA-treated cells , were treated with VepA . For lysosomes from ATP6V0C-silenced cells , CatD release was partially decreased ( Figure 5N ) , indicating that the knockdown of ATP6V0C reduces VepA-induced lysosomal rupture . Thus , these results indicate that ATP6V0C is involved in VepA-induced lysosomal rupture .
Pathogens manipulate host cell death to facilitate their ability to cause infections [31] , [32] . The bacterial pathogen V . parahaemolyticus elicits T3SS1-dependent non-apoptotic and caspase-independent cell death during infection ( Figure S1 ) [11] , [12] . Among the effectors translocated by the T3SS1 of V . parahaemolyticus , VepA is thought to play an important role in V . parahaemolyticus-induced cytotoxicity because the ΔvepA strain has lost the majority of the cytotoxicity of the wild-type [14] , [15] . In this study , we showed that VepA itself is cytotoxic and acts only inside cells , not outside cells ( Figure 1 ) . It is reasonable for VepA to be cytotoxic only inside cells because it is injected into the cytoplasm of host cells by T3SS . Among the vibrios , V . alginolyticus , V . harveyi and V . tubiashii possess T3SSs that are highly homologous to V . parahaemolyticus T3SS1 [8] . The V . alginolyticus T3SS exhibits cytotoxicity in mammalian and fish cells [33] , [34] . Because VepA is also conserved among these species , it is possible that VepA homologs are involved in the cytotoxicity of other Vibrio species . Our yeast genome-wide screen revealed that subunit c of V-ATPase is indispensable for the toxicity of VepA in yeast ( Figure 2 ) . In HeLa cells , knockdown of ATP6V0C with siRNA decreased VepA-mediated cytotoxicity significantly but not completely ( Figure 3 ) , presumably due to the insufficient knockdown efficiency of ATP6V0C , as validated in Figure 3B . The expression of GFP-Vc1 , a cytoplasmic loop of ATP6V0C , also reduced the cytotoxicity significantly but only partially ( Figure S5B ) . GFP-fused cytoplasmic loops exhibited a diffuse cytoplasmic pattern and no lysosomal localization ( data not shown ) , which may result in insufficient competition with ATP6V0C . Our pull-down assays suggested that VepA prefers Vc1 , but Vc2 was also weakly associated with VepA ( Figure S5C ) . An alternative explanation for the residual cytotoxicity is that VepA may recognize not only the cytoplasmic loop 1 of ATP6V0C but also the more complicated structure of ATP6V0C . Thus , although we cannot exclude the possibility that VepA has other unknown cellular targets that may also be involved in cytotoxicity at this stage , VepA-mediated cytotoxicity in HeLa cells at least partially requires subunit c of V-ATPase , supporting the validity of our yeast genome-wide screen . V-ATPase is composed of multisubunits , all of which are required for the proton transport activity . The V-ATPase complex is divided into the two domains , V1 and V0 domains , which can assemble independently [22] . Subunit c of V-ATPase , a component of the V0 domain , is also known as ductin , which is thought to form a channel by itself [24] . Therefore , even in the absence of other ATPase components , it is possible that molecular complexes containing subunit c of V-ATPase are expressed in the membrane , although these complexes may not function as ATPases . In yeast , in the assembly of the V0 domain , deletions in the V0 subunits have been reported to result in the failure of the V0 domain to localize to vacuolar membranes [35] , [36] . However , in our yeast toxicity assays , presented in Figure 2 , VepA was shown to be toxic to the V0 subunit mutants except for Δvma3 . Therefore , although it should be noted that VepA was ectopically overexpressed in yeast in these experiments , we cannot completely exclude the possibility that VepA may target subunit c of V-ATPase , which is also located in compartments other than vacuolar membranes . This possibility will be explored in the future . Biochemical cell fractionation revealed the distribution of VepA is almost identical to that of V-ATPases ( Figure S3 ) . V-ATPases are expressed in acidic organelles and at plasma membranes [22] . Indeed , the distribution of V-ATPases was similar to that of lysosomes and plasma membranes in Figure S3B . By contrast , the distribution of VepAΔC , which did not interact with ATP6V0C , was different from that of wild-type VepA , suggesting that VepA is associated with these organelles via the interaction with subunit c of V-ATPases . V-ATPases are evolutionally conserved among eukaryotic cells [23] . A bacterial virulence factor that targets such a broadly distributed molecule among species is an efficient way to ensure a wide range of host species susceptibility . The Pseudomonas aeruginosa pigment pyocyanin and Legionella pneumophila SidK inactivate V-ATPase [37] , [38] . Pyocyanin has been implicated in chronic P . aeruginosa infection in cystic fibrosis [39] . However , some yeast V-ATPase mutant strains are more sensitive to pyocyanin than the wild-type strain , a property that is distinct from the toxicity of VepA , which is not toxic to Δvma3 . SidK , a type IV secreted effector , targets subunit A of V-ATPase , which contains the ATP hydrolytic site , and inhibits ATP hydrolysis to prevent the proton pump function of V-ATPase and phagosomal acidification [38] . By contrast , our data indicate that , although VepA targeted subunit c of V-ATPase , the function of V-ATPase is dispensable for the toxicity of VepA ( Figure 2 and Figure 3 ) . Unlike VepA , which showed severe and acute cytotoxic activity within 4 h after infection ( Figure 1 and S6C ) , SidK appeared not to be cytotoxic during the early infection period because macrophages loaded with the SidK protein survive longer than 24 h [38] . This difference may reflect the difference between the infection strategies of the two pathogens: V . parahaemolyticus is an extracellular pathogen that causes acute infection [5] , whereas L . pneumophila is an intracellular pathogen that survives and replicates in phagosomes and therefore needs to avoid phagosomal acidification [40] . Infection with V . parahaemolyticus caused VepA injection-dependent leakage of the lysosomal contents ( Figure 5 ) . Lysosomes are described as “suicide bags” , because they contain numerous hydrolases [41] . Lysosomotropic agents such as H2O2 and sphingosine induce lysosomal membrane permeabilization and subsequent cell death [42] . Lysosomal rupture-induced cell death is thought to be dependent on the extent of lysosomal damage; partial and moderate rupture causes apoptosis , whereas more severe damage leads to necrosis [25] , [26] . V . parahaemolyticus infection-mediated leakage of lysosomal contents was observed not only with a small-molecule dye AO but also with the protein CatD , suggesting that extensive lysosomal membrane permeabilization occurs within hours of infection . Notably , our cell-free assay showed that VepA is sufficient to reproduce infection-mediated lysosomal rupture , whereas VepAΔC , which is deficient for the association with ATP6V0C , could not induce the leakage of lysosomal contents . Moreover , the knockdown of ATP6V0C reduced VepA-induced lysosomal rupture ( Figure 5L and 5N ) , indicating that ATP6V0C is required of VepA-induced lysosomal rupture . Thus , our data provide the first example of a bacterial T3SS effector that ruptures lysosomes directly to induce cell death . Various bacterial effectors are known to have enzymatic activities such as proteolytic processing and post-translational modification [7] . However , despite experimental efforts , we could not detect the processing or modification of ATP6V0C by VepA in this study . Although it is not yet known how the interaction of VepA with ATP6V0C leads to destabilization of lysosomal membranes , ATP6V0C may serve as a scaffold for VepA , facilitating the access of VepA to lysosomal membranes and subsequent lysosomal rupture . Alternatively , the association of VepA with ATP6V0C may destabilize the ATP6V0C complex , which may result in lysosomal destabilization . Two lysosome stabilizers , U18666A and DFO , partially inhibited T3SS1-dependent cytotoxicity ( Figure S6E and S6F ) . U18666A inhibits transport of cholesterol from lysosomes to the ER , which causes the accumulation of cholesterol in lysosomes [29] . DFO is known to protect against H2O2-induced lysosome rupture by chelating intralysosomal iron [30] . In contrast to those direct effects on lysosomal stabilization , the incubation of heat shock protein 70 ( Hsp70 ) with cells , which is reported to stabilize lysosomes by an indirect effect that stimulates acid sphingomyelinase activity [43] , failed to prevent T3SS1-dependent cytotoxicity ( data not shown ) . Thus , these results suggest that VepA physically destabilizes lysosomal membranes . The cell death induced by V . parahaemolyticus has also been reported to be associated with autophagy [11] . Although autophagy is well known to promote cell survival in response to various stimuli [44] , recent studies have indicated that autophagy also plays a role as an executor of cell death in some aspects [45] , [46] . However , the role of autophagy induced by V . parahaemolyticus in cell death is unclear . A previous report showed that V . parahaemolyticus does not activate autophagy in Atg5-/- murine embryonic fibroblasts , indicating that V . parahaemolyticus-induced autophagy is dependent on ATG5 [16] . The results of this study were consistent with those of a prior study , demonstrating that ATG5 depletion by siRNA inhibited V . parahaemolyticus-induced autophagy in HeLa cells ( Figure S1E ) . However , ATG5 depletion did not affect T3SS1-dependent cytotoxicity ( Figure S1C ) . Moreover , Δatg5 and Δatg8 yeast strains , both of which are deficient in the autophagic process , were not resistant to VepA ( Figure S2B ) . Indeed , although there is a link between autophagy and lysosomal biogenesis [47] , our results indicate that autophagy does not contribute to the cell death induced by V . parahaemolyticus . Thus , V . parahaemolyticus induces “cell death with autophagy” but not “cell death by autophagy” [48] . It is thus possible that autophagy may play a protective role against cytotoxicity of V . parahaemolyticus . It has also been reported that VepA is linked to the activation of MAPK signaling when cells are infected with V . parahaemolyticus [19] . MAPK cascades are activated by various stimuli including cellular stress [49] . A recent report has shown that deficiency in the tumor susceptibility gene 101 leads to lysosomal distention as well as induction of autophagy and MAPK activation [50] . Thus , it is possible that lysosomal stress may be linked to induction of stress response , such as activation of MAPK and induction of autophagy . In conclusion , we demonstrated that VepA targets ATP6V0C and ruptures lysosomes . Although the mechanism of lysosomal rupture by VepA warrants further exploration , it is intriguing that a pathogenic bacterium induces cell death by causing the host cell to leak its own dangerous content , which is akin to striking the Achilles' heel of the host cell .
The Vibrio parahaemolyticus strains POR-3 ( RIMD2210633ΔtdhASΔvcrD2 , which is T3SS2 deficient ) and POR-3ΔvcrD1 ( which lacks both T3SS1 and T3SS2 ) were described previously [9] , [14] . The POR-3ΔvepA strain was created as described previously [14] . To complement the mutant with vepA or vepAΔC ( 1–400 ) , the pSA-19CS-MCS vector was used as described previously [14] . Saccharomyces cerevisiae BY4730 ( MATa leu2Δ0 met15Δ0 ura3Δ0 ) was obtained from Open Biosystems . The plasmids and oligonucleotide primers used in this study are listed in Table S1 . Bafilomycin A and concanamycin A ( both used at 100 nM ) were purchased from Sigma . Acridine orange was purchased from Invitrogen . U18666A , deferoxamine , SP600125 , U0126 and SB20358 were purchased from Sigma . The pan-caspase inhibitor Z-VAD-FMK was from Medical & Biological Laboratories . The following antibodies were used: antibodies against Lamp-1 , CD49b , Grb78 , GM130 , nucleoporin-62 , Bcl-2 and Hsp90 ( BD Biosciences ) ; antibodies against β-actin , Flag and poly-histidine ( Sigma ) ; anti-ATP6V1A ( Abnova ) ; anti-ATP6V0D1 ( Abcam ) ; anti-CatD ( Cell Signaling Technology ) ; HRP-conjugated streptavidin ( Pierce ) ; anti-ATP6V0C ( Millipore ) ; HRP-conjugated anti-GFP ( Miltenyi Biotech ) ; and anti-ATG5 and anti-LC3 ( Medical & Biological Laboratories ) . The anti-VepA antibody has been described elsewhere [15] . HeLa , RAW264 . 7 and 293T cells were maintained in DMEM ( Sigma ) containing 10% FBS ( Sigma ) at 37°C in 5% CO2 . For infections , bacteria were used to challenge HeLa cells at a multiplicity of infection of 10 [10] . The cytotoxicity assay was performed using the CytoTox96 Non-Radioactive Cytotoxicity Assay Kit ( Promega ) as previously described [9] . Yeast was transformed using the Frozen-EX Transformation Kit II ( Zymo Research ) . Transformants were grown on SC plates lacking uracil and containing 2% glucose ( SC-Ura+Glc ) at 30°C for 48–72 h . Growing colonies were picked and cultured in SD media lacking uracil and containing 2% glucose ( SD-Ura+Glc ) . Yeast was washed once with 0 . 67% yeast nitrogen broth without amino acids and adjusted to an optical density of 1 at 600 nm . SC plates lacking uracil and containing glucose or galactose ( SC-Ura+Gal ) were spotted with 5-µl aliquots of 10-fold serial dilutions and then incubated at 30°C for 72 h . For the complementation of VMA3 in Δvma3 , a LEU2 plasmid pRS415 , encoding VMA3 along with 500 bp upstream and downstream of VMA3 was used . His-tagged VepA or VepAΔC protein was purified as described previously [15] . To construct pATP6V0C-Flag , cDNA for ATP6V0C with a Flag-tag and a stop codon at the C-terminus was inserted into pEGFP-N1 , yielding the ATP6V0C-Flag construct without GFP . For DNA transfection , 293T cells were used . To transfect the VepA constructs , 293T cells were seeded on collagen-coated coverslips in 6-well plates and grown for 24 h . The cells were then transfected with pEGFP-C1 , pEGFP-VepA or pEGFP-VepAΔC using Lipofectamine 2000 ( Invitrogen ) . The cells were fixed with 3% paraformaldehyde and permeabilized with 0 . 2% Triton X-100 ( TX-100 ) 24 h post-transfection and then stained with rhodamine-phalloidin . The coverslips were analyzed by fluorescence microscopy using a Biozero BZ-8100 microscope ( Keyence ) . To infect cells expressing the ATP6V0C constructs , 293T cells seeded in 96-well plates were transfected with pEGFP-C1 , pEGFP-Vc1 or pEGFP-Vc2 using Lipofectamine LTX ( Invitrogen ) . After 24 h , the cells were infected with POR-3 at a multiplicity of infection of 10 for 3 h and then assessed using the cytotoxicity assay . Proteins were introduced into cells using GenomONE-Neo ( Ishihara Sangyo ) according to the manufacturer's instructions . Briefly , 10 µl of HVJ-envelope was mixed with 2 . 5 µl of Reagent A and incubated for 5 min on ice . The suspension was mixed with 5 µg of protein and then with 1 . 5 µl of Reagent B . After centrifugation at 10 , 000 g for 5 min , the supernatant was removed . The pellets were resuspended in 15 µl of Reagent Buffer , followed by the addition of 2 . 5 µl of Reagent C to complete the preparation of the HVJ-liposome-including proteins . A one-eighth aliquot of the envelope was added to cells in a 96-well plate . The cells were incubated at 37°C for 1 . 5 h and then washed with phosphate-buffered saline ( PBS ) . The medium was replaced with fresh medium , and the cells were incubated at 37°C for 2 . 5 h , and cytotoxicity was determined by assaying the cellular dehydrogenase activity using Cell Counting Kit-8 ( Dojindo ) . Yeast MATa haploid non-essential gene knockout strains from Open Biosystems [21] were grown in YPD medium in 96-well plates at 30°C . All of the strains from a single plate were then pooled into a single culture , and transformed with p426-VepA as described above . Transformants were plated on SC-Ura+Gal and incubated at 30°C for 3 days . To determine the plating efficiency , transformants were also plated on SC-Ura+Glc . Colonies grown on galactose plates were picked and incubated in SD-Ura+Glc . Chromosomal DNA was purified by phenol/chloroform extraction with glass beads as described elsewhere [51] . Barcode-tag sequence regions of the extracted yeast DNA were amplified by polymerase chain reaction . Then , amplicons were purified using ExoSAP-IT ( GE healthcare ) and sequenced to verify the barcode-tag . We eliminated mutants that were only recovered in a single clone to exclude false positives . For the second selection , the identified mutants were individually transformed with p426-VepA , and their growth on SC-Ura+Gal was examined . A flowchart summarizing the protocol for the yeast genome-wide screen used in this study is shown in Figure 2B . Ambion's Silencer Select Validated siRNAs and Silencer Select Pre-designed siRNAs were used to knockdown ATP6V0C and ATP6V1A , respectively . HeLa cells were reverse-transfected with 20 nM Silencer Negative Control#2 siRNA or with two independent siRNAs targeting ATP6V0C ( V0C#1; s80 , V0C#2; s81 ) or ATP6V1A ( V1A#1; s1791 , V1A#2; s1792 ) using the siPORT NeoFX Transfection Agent ( Ambion ) . Then , 96 h after siRNA transfection , the cells were infected with V . parahaemolyticus strain POR-3 as described above . To determine the knockdown efficiency , immunoblot analysis was performed using anti-ATP6V1A and anti-ATP6V0C antibodies . To deplete ATG5 , FlexiTube siRNAs ( Qiagen ) were used . HeLa cells were transfected with AllStars Negative Control siRNAs or two separate siRNAs targeting ATG5 ( ATG5_2; #1 , ATG5_6; #2 ) using HiPerFect Transfection Reagent ( Qiagen ) . Then , 72 h post-transfection , the cells were infected with strain POR-3 for 4 h , and cytotoxicity was evaluated . For pull-down assays , 293T cells expressing ATP6V0C-Flag or RAW264 . 7 cells were lysed with RIPA buffer containing a protease inhibitor cocktail ( Sigma ) . VepA and VepAΔC were biotinylated using EZ-Link Sulfo-NHS-SS-Biotinylation kits ( Pierce ) . Biotinylated proteins were captured with Streptavidin Sepharose beads ( GE healthcare ) . Then , VepA- or VepAΔC-immobilized beads were incubated with cell lysates on a rotary shaker for 2 h at 4°C . The beads were washed with RIPA buffer five times , and resuspended in SDS-loading buffer to elute bound proteins from the beads . Eluates were subjected to SDS-PAGE and immunoblot analysis or LC-MS/MS analysis as described previously [52] . Proteins were identified by a database search using Mascot ( Matrix Science ) . Subcellular fractionation was performed using the ProteoExtract Subcellular Proteome Extraction Kit ( Calbiochem ) according to the manufacturer's instructions . For organelle fractionations , we used the Lysosome Enrichment Kit for Tissue and Cultured Cells ( Pierce ) according to the manufacturer's instructions . Briefly , the lysates from infected cells were applied to a 15–30% OptiPrep density gradient . After ultra-centrifugation at 150 , 000 g for 2 h at 4°C in a SW55 rotor ( Beckman Coulter ) , 800-µl fractions were collected from the top of the tube . Each fraction was diluted with PBS and centrifuged at 20 , 000 g for 30 min at 4°C . After removing the supernatants , the pellets were dissolved with SDS-loading buffer and subjected to SDS-PAGE and immunoblot analysis . To prepare the cytosolic fraction , HeLa cells were solubilized with saponin buffer ( 0 . 1% saponin and protease inhibitors in PBS ) for 15 min on ice . After centrifugation at 15 , 000 g for 10 min , the supernatants were collected . The cytosolic fraction was free of membranes as verified by immunoblots using an anti-Lamp-1 antibody . Lysis refers to treating the cells with 1% TX-100 to induce the total release of CatD as a positive control . HeLa cells grown on 35-mm glass bottom dishes were treated with 2 µg ml−1 AO at 37°C for 15 min . After washing with PBS , the medium was replaced with fresh medium , followed by infection or exposure to TDH . Fluorescence micrographs of AO-stained cells were obtained using the same fluorescence intensity and exposure time . The intensity of green fluorescence was quantitated with a BZ Analyzer II ( Keyence ) . The cell-free lysosome-enriched fraction was prepared as described [53] . Briefly , HeLa cells were homogenized in ice-cold TS buffer ( 10 mM Tris-HCl , pH 7 . 5 , 250 mM sucrose and protease inhibitors ) . Then , the cells were centrifuged at 1 , 000 g to pellet the nuclei and cell debris . The supernatants were centrifuged at 3 , 000 g . The supernatants were further centrifuged at 17 , 000 g , and the resulting pellets , which contained the lysosomes , were resuspended with PBS and used as a cell-free lysosomal preparation . Then , 50-µl aliquots of prepared lysosomes ( 50 µg ) were exposed to 10 µg ml−1 VepA or VepAΔC for 2 h at 37°C . The reaction mixture was centrifuged at 20 , 000 g for 30 min at 4°C to separate the pellets , which contained the lysosomes , from the supernatants , which contained the material that leaked from the lysosomes . Both fractions were subjected to SDS-PAGE and immunoblot analyses using anti-CatD and anti-Lamp-1 antibodies . To induce autophagy , HeLa cells were starved in Earle's balanced salt solution for 6 h or infected with V . parahaemolyticus POR-3 or ΔvepA for 3 h . For the nutrient-rich condition , cells were cultured in DMEM containing 10% FBS . To measure autophagy , the conversion of LC3-I to LC3-II was monitored by immunoblotting as previously described [54] . Statistical analysis was performed using the two-tailed Student's t-test . | Vibrio parahaemolyticus is a bacterial pathogen that causes food-borne gastroenteritis and also wound infection and septicemia . It exhibits cytotoxicity that is dependent on its type III secretion system ( T3SS1 ) during the infection of mammalian cells . Although an effector VepA plays a major role in the cytotoxicity , the mechanism was unknown . Here we show that VepA targets subunit c of the vacuolar H+-ATPase ( V-ATPase ) and induces the rupture of host cell lysosomes . We found that VepA alone is cytotoxic in HeLa cells and also toxic in yeast Saccharomyces cerevisiae . Using a yeast genome-wide screening , we identified yeast V-ATPase subunit c as essential for the toxicity of VepA to yeast . We also demonstrated that knockdown of V-ATPase subunit c decreased VepA-mediated cytotoxicity toward HeLa cells and that VepA interacted with subunit c of V-ATPase . During infection , lysosomal contents leaked into the cytosol prior to cell lysis , and VepA was necessary and sufficient for this leakage . Our data suggest that a bacterial effector VepA ruptures lysosomes , the “suicide bags” of host cells , by targeting the evolutionarily conserved V-ATPase , and elicits subsequent cell death . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"death",
"gram",
"negative",
"molecular",
"cell",
"biology",
"cell",
"biology",
"genetic",
"screens",
"genetics",
"microbial",
"pathogens",
"genetics",
"and",
"genomics",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens"
] | 2012 | A Cytotoxic Type III Secretion Effector of Vibrio parahaemolyticus Targets Vacuolar H+-ATPase Subunit c and Ruptures Host Cell Lysosomes |
Previous explanations of computations performed by recurrent networks have focused on symmetrically connected saturating neurons and their convergence toward attractors . Here we analyze the behavior of asymmetrical connected networks of linear threshold neurons , whose positive response is unbounded . We show that , for a wide range of parameters , this asymmetry brings interesting and computationally useful dynamical properties . When driven by input , the network explores potential solutions through highly unstable ‘expansion’ dynamics . This expansion is steered and constrained by negative divergence of the dynamics , which ensures that the dimensionality of the solution space continues to reduce until an acceptable solution manifold is reached . Then the system contracts stably on this manifold towards its final solution trajectory . The unstable positive feedback and cross inhibition that underlie expansion and divergence are common motifs in molecular and neuronal networks . Therefore we propose that very simple organizational constraints that combine these motifs can lead to spontaneous computation and so to the spontaneous modification of entropy that is characteristic of living systems .
The principles of biological computation are not well understood . Although the Turing Machine and related concepts [1–3] have provided powerful models for understanding and developing technological computing , they have provided less insight for biological computation because they generally assume that the machines themselves , as well as their initial program and data are granted as input . In contrast , the organization of states and transitions of the biological process arise out of phylogenetic and ontogenetic configuration processes and execute autonomously without the intervention of an intelligent external programmer and controller being necessary to supply already encoded organizationally relevant information . Our goal here is to make steps towards understanding biological computation [4–6] , by considering the behavior of a simple non-linear dynamical system composed of asymmetrically inter-connected linear-threshold neurons . We suppose that such computations entail a mapping from some input towards a limited ( low entropy ) region of phase space , which is the solution [7] . We do not suppose that the computational goal is known—only that computation must conform to this basic entropy reducing process . Here we describe the organizational constraints that make such spontaneous computation possible . Previous authors have explained neural network computation in terms of the convergence of special dynamical systems , and emphasized the attractors to which they converge [8–13] . For example , Hopfield [9 , 10] has shown how and why the dynamics of symmetrically connected neurons with saturating outputs converge to attractor states; and others have offered similar insights for symmetrically connected linear threshold neurons [14–16] . However , interactions between inhibitory and excitatory neurons are clearly asymmetric , making these studies ill suited to study biological computation . To the extent that asymmetrical networks have been considered at all , this has been through assumptions that reduce asymmetrical networks to approximate symmetry . By contrast , we consider here the dynamics of fully asymmetrical networks , and discover that asymmetry contributes strongly to computational behavior . The scope of our work is restricted to recurrent neural networks with asymmetric coupling that express an important and ubiquitous behavior: soft winner-take-all ( sWTA ) dynamics . We present a formal account of the response of these networks to exogenous perturbations and use a form of non-linear stability analysis ( contraction analysis [17] ) to characterize the itinerant transients than ensue , and which have useful interpretations in terms of neuronal computation and information theory . Contraction Theory offers a more flexible framework than the conventional Lyapunov approach to non-linear stability ( See Methods for details ) . This is particularly the case for non-autonomous systems such as our network , in which external inputs can vary with time . We explore particularly the behavior of network computation during the non-equilibrium phase , when the network is traversing its state-space seeking for a solution . We show that the ability of the network to explore potential solutions depends on highly unstable ’expansion’ dynamics driven by recurrent excitation . This expansion is steered and constrained by negative divergence of the dynamics , which ensures that the dimensionality of the solution space continues to reduce until an acceptable solution manifold is reached . The system then ’contracts’ stably on this manifold [17] towards its final solution trajectory , which is not necessarily converging to a fixed point . We argue that the simple principle of unstable expansion constrained by negative divergence provides the central organizing drive for more general autonomous biological systems from molecular networks , through neurons , to society . Consider a simple network of non-linear neuron-like elements whose task it is to compute the solution to some problem . The states of the computation are encoded in the activations ( firing rates ) of the neurons , and the computational transitions between these states arise out of their synaptic interactions . The overall trajectory resulting from the successive transitions through its state space express its computation [18–20] . In current technological systems the hardware states of the system are encoded on binary nodes whose discrete states are imposed by signal restoring [21] circuitry [19] . This signal restoration is achieved by extremely high gain , so that a small input bias will drive the node into saturation at one of its two voltage limits . Biology rarely commands such sharply demarcated states and transitions . Instead , molecular and electrophysiological activation functions are often approximately sigmoidal ( eg Hill functions , voltage dependent conductance , neuronal current-discharge curves , etc ) . However , neuronal systems do not typically run in saturation . The typical activation of a neuron is thresholded below , and above this threshold it makes use of only the lower part of its dynamic range . It very rarely enters saturation at the upper end of its activation range . Therefore , a suitable model for neuronal activation is a thresholded linear one . That is , their activity is bounded from below , but their positive activity is essentially unbounded ( over any practical range of discharge ) . This is a very well studied model [14–16 , 22] . As in our previous work , the neuronal network model is composed of thresholded linear neuron-like units coupled through positive ( excitatory ) and negative ( inhibitory ) connections ( see Fig . 1a ) . The unbounded positive range of neuron activation implies that the global stability of networks of these neurons must arise out of their collective interactions rather than from saturation of their individual activation functions as assumed by for example [9 , 10] . The key interaction here is the inhibitory feedback , which must at least ensure that not all neurons can simultaneously increase their activation [16] . Previous studies of such models have focused on the mathematically more tractable case in which the connections between neurons are symmetrical , and have no transmission delays . Our networks , by contrast , need not be symmetrical and may have transmission delays . Indeed , the asymmetry of connections will be used to computational advantage , not offered by symmetrical networks .
The network contains two fundamental circuit motifs ( Fig . 1A ) : excitatory neurons that project both onto themselves and others , and inhibitory neurons which receive excitatory input from the same neurons that they inhibit . Several instances of these motifs together compose the general recurrent circuit that we investigate . In its simplest form , this recurrent network has the following form . There are N neurons , N-1 of which are excitatory and one ( index N ) is inhibitory ( Fig . 1B , C ) . The excitatory neurons xi ≠ N receive an optional external input Ii , and excitatory feedback α from themself and nearby excitatory neurons ( α1 and α2 , respectively ) . The single inhibitory neuron xn sums the β2 weighted input from all the excitatory neurons , and sends a common β1 inhibitory signal to each of its excitatory neurons . Each neuron has a resistive constant leak term Gi . The dynamics of this simple network are: τ x ˙ i + G i x i = f ( I i ( t ) + α 1 x i + α 2 x i - 1 + α 2 x i + 1 - β 1 x N ) ( 1 ) τ x ˙ N + G i x N = f ( β 2 ∑ j = 1 N - 1 x j ) ( 2 ) where f ( x ) is a non-saturating rectification non-linearity . Here , we take f ( x ) = max ( x , 0 ) , making our neurons linear threshold neurons ( LTNs ) . As long as the key parameters ( αi , βi ) satisfy rather broad constraints [23] , this network behaves as a soft winner-take-all by allowing only a small number of active neurons to emerge , and that solution depends on the particular input pattern I ( t ) . After convergence to the solution , the active neurons ( winners ) will express an amplified version of the input I ( t ) , that is they remain sensitive to the input even if it changes . After convergence to steady state in response to constant input Ii > 0 , the activation of the winner i is: x i = I i 1 - α 1 + β 1 β 2 = g I i ( 3 ) where g = 1 1 − α 1 + β 1 β 2 is the gain of the network . Importantly , this gain can be g > > 1 due to excitatory recurrence , which amplifies the signal in a manner controlled by the feedback loop . While above assumes α2 = 0 , similar arguments hold without this assumption [23] . The dynamics of the non-linear system are τ x ̇ = f ( Wx + I ( t ) ) − Gx ( where f applies the scalar function f ( x ) component-wise ) . For example , a minimal WTA with two excitatory and one inhibitory units , x = [x1 , x2 , x3] ( Fig . 1B , C ) , the weight matrix W is W=[ α 1 0 − β 1 0 α 1 − β 1 β 2 β 2 0 ] ( 4 ) where G = diag ( G1 , … , Gn ) is a diagonal matrix containing the dissipative leak terms for each unit . The time constant of the system is τ G . Now we consider the conditions under which the non-linear system is guaranteed to be stable but at the same time powerful , i . e . permitting high gain . Contraction Theory assesses the stability of non-linear systems x˙=f ( x , t ) using virtual displacements of its state at any point with respect to a chosen uniformly positive definite metric M ( x , t ) , obtained by a transformation Θ ( x , t ) ( see Methods ) . The key Theorem 2 of [17] asserts that if the temporal evolution of any virtual displacement in this metric space tends to zero , then all other displacement trajectories in this space will also contract ( shrink ) to the same ( common ) trajectory . For our present case , this theorem implies ( see Methods ) that the trajectories of a system of neurons with Jacobian J are exponentially contracting if Θ J Θ - 1 < 0 ( 5 ) The Jacobian J has dimension N and describes the connection matrix of the network and Θ is a transformation matrix that provides a suitable choice of coordinates . For a WTA with weight matrix W , the Jacobian is J = ∂ f ∂ x = Σ W − G ( 6 ) where Σ = diag ( σ1 , … , σn ) is a switching matrix , whose diagonal elements σi ∈ 0 , 1 indicate whether unit i is currently active or not . Following [16] we will call the subset of N neurons that are currently above threshold the active set . Those that are inactive cannot contribute to the dynamics of the network . Thus , we may distinguish between the static anatomical connectivity W of the network , and the dynamic functional connectivity that involves only the subset of currently active neurons . This active set is described by the switch matrix Σ . The Jacobian expresses which units contribute to the dynamics at any point of time , and is thus a reflection of functional rather than anatomical connectivity . Each possible effective weight matrix has a corresponding effective Jacobian . Thus , ΣW − G is the effective ( functional ) Jacobian of the network , in which only active units have an influence on the dynamics . The active sets may be either stable or unstable . In previous work on the hybrid analog and digital nature of processing by symmetric networks , Hahnloser et al . refer to such sets as active and forbidden , respectively [16] . They further show that the eigenvectors associated with positive eigenvalues of forbidden sets are mixed . That is , the eigenvector contains at least one component whose sign is opposite to the remainder of the components . This property ensures that a neuron will finally fall beneath threshold , and that the composition of the active set must change . We now generalize this concept to our asymmetrical networks , and refer to permitted and forbidden subspaces rather than sets , to emphasize the changes in the space in which the computational dynamics play out . It is the instability ( exponential divergence of neighboring trajectories ) of the forbidden subspaces rather than stability that drives the computational process . This instability can be approached also through Theorem 2 of [17] , which notes that if the minimal eigenvalue of the symmetric part of the Jacobian is strictly positive , then it follows that two neighboring trajectories will diverge exponentially . We will use ’expanding’ to refer to this unstable , exponentially diverging behavior of a set of neurons to avoid confusion with Gaussian divergence , which we will need to invoke in a different context , below . Thus , we will say that the dynamics of a set of active neurons is expanding if Θ V J V T Θ − 1 > 0 ( 7 ) where V is a projection matrix which describes subspaces that are unstable . For example , for a circuit with two excitatory units that cannot both be simultaneously active , V is V = α 0 − β 1 0 α − β 1 ( 8 ) and Θ is a metric . The constraint ( 7 ) asserts that the system escapes the unstable subspaces where Vx is constant . This guarantees that For V as defined above Vx = [ α x 1 − β 1 x 3 α x 2 − β 1 x 3 ] . Each row represents one excitatory unit . Guaranteeing that Vx cannot remain constant for a particular subset implements the requirement that for a subspace to be forbidden , it cannot be a steady state because if it were Vx would remain constant after convergence . The parameter conditions under which Eqn 7 holds , are given in detail in [24] . Importantly , these conditions guarantee that when the dynamics of our asymmetric network occupies an unstable subspace , all eigenvectors are mixed ( see Methods for proof ) . Consequently , as in the symmetric case of [16 , 25] , this unstable subspace will be left ( it is forbidden ) , because one unit will fall beneath its threshold exponentially quickly and so become inactive . The dynamics of our asymmetric networks can now be explained in terms of the contraction theory framework outlined above . Consider a simple network consisting of N = 5 neurons , one of which is inhibitory and enforces competition through shared inhibition ( Fig . 2A ) . For suitable parameters that satisfy the contraction constraints ( see [24] ) , this network will contract towards a steady state for any input . The steady state will be such that the network amplifies one of the inputs while suppressing all others ( for example , Fig . 2B ) . During this process of output selection and amplification , the network passes through a sequence of transformations , at each of which a different subset of units becomes active whereas the remainder are driven beneath threshold and so are inactive . These transformations continue while the network is in its expansion phase and cease when the network contracts . The network is in the expansion phase while the effective Jacobian is positive definite , and contracts when the Jacobian becomes negative definite ( Fig . 2C ) . The computational process of selecting a solution , conditional on inputs and synaptic constraints , involves testing successive subspaces , expressed as changing patterns of neuronal activation ( Fig . 2B ) . Subspaces do not offer a solution ( are forbidden ) if for that subspace VJVT is positive definite . In this case its dynamics are expanding , and because all eigenvectors in this subspace are mixed , the subspace will finally switch . Subspaces that offer solutions ( are permitted ) will have an effective Jacobian that is negative definite in some metric ( they are contracting ) , and so the system will not leave such a subspace provided that the input remains fixed . Note that there can be subspaces whose Jacobian is neither positive nor negative definite , which are then neither permitted or forbidden . However , by definition , the networks we describe assure that each possible subspace is either permitted or forbidden . Consider the case in which the computational process begins in a forbidden subspace ( Fig . 2F ) . The process then passes successively through several forbidden subspaces before reaching a permitted subspace . Two properties ensure this remarkable orderly progression towards a permitted subspace . Firstly , the process is driven by the instability that ensures that forbidden subspaces are left . However the trajectory is associated with a progressive reduction in the maximum positive eigenvalue of the active set ( Fig . 2C ) . Secondly , an orderly progression through forbidden subspaces is ensured by systematic reduction of the state space through Gaussian divergence . Gauss’ theorem d d t δ V = div ( d dt δ z ) δ V ( 9 ) asserts that , in the absence of random fluctuations , any volume element δV shrinks exponentially to zero for uniformly negative definite div ( d dt δ z ) . This implies convergence to an ( n−1 ) dimensional manifold rather than to a single trajectory . For our system , the Gaussian divergence is the trace of the Jacobian J or equivalently the sum of all of its eigenvalues . Note that this is a much weaker requirement than full contraction . In particular , we are concerned about the trace of the effective Jacobian , which is the Jacobian of only the active elements , because the inactive elements ( those below threshold ) do not contribute to the dynamics . The divergence quantifies the rate at which the volume shrinks ( exponential ) in its n dimensional manifold towards a ( n−1 ) dimensional manifold . The system will enter the ( n−1 ) dimensional and continue its evolution in this new active subset , and so on , until the reduced dimensional system finally enters a permitted subspace ( which is contracting ) . Note that here we refer to the dimensionality of the system dynamics . This dimensionality is not necessarily equal to the number of active units ( i . e . at steady state or when one of the units is constant ) . Note that negative Gaussian divergence does not follow automatically from expansion . In fact most expanding sets will not have negative Gaussian divergence , and so will not be forbidden For example consider the linear system x . = Wx with W = [ 1 0 0 0 1 0 0 0 1 ] . This system is expanding but it does not have negative divergence . For orderly computation to proceed , such sets must not exist . We require that all forbidden subspaces are expanding as defined by contraction theory as well as have negative Gaussian divergence . Indeed , for our LTN networks all forbidden subspaces are guaranteed to satisfy both these properties . Because their positive output is unbounded , linear thresholded neurons are essentially insensitive to the range of their positive inputs . Networks of these neurons amplify their inputs in a scale-free manner , according to the slope of their activation functions and the network gain induced by their connections . As explained above , this high gain drives the exploration and selection dynamics of the computational process . The network harnesses its high gain to steer the computation through its asymmetric connections . Indeed , the asymmetric nature of the connectivity in our network is central to its operation , and not a minor deviation from symmetry ( approximate symmetry , [26] ) . The interplay between high-gain and steering can be appreciated by considering the behavior of the system within one of the subspaces of its computation . Consider a linear system of the form x . = f ( x , t ) , which can be written as x . = M x + u where M is a matrix of connection weights and u a vector of constant inputs . For the example of a WTA with 2 excitatory units and 1 inhibitory unit , M = l 1 α − G 0 − l 1 β 1 0 l 2 α − G − l 2 β 1 l 3 β 2 l 3 β 2 − G ( 10 ) Setting M = M1+M2 and defining M 1 = l 1 α − G 0 0 0 l 2 α − G 0 0 0 − G ( 11 ) M 2 = 0 0 − l 1 β 1 0 0 − l 2 β 1 l 3 β 2 l 3 β 2 0 ( 12 ) provides a decomposition into a component with negative divergence and zero divergence ( see methods for an example ) . Any vector field f ( x ) can be written as the sum of a gradient field ∇V ( x ) and a rotational vector field ρ ( x ) , i . e . a vector field whose divergence is zero . In analogy , f ( x ) = x . = M 1 x + u + M 2 x where ∇V ( x ) = M1x+u and ρ ( x ) = M2x . For our network M1 is the expansion/contraction component and M2 is the rotational component . This is an application of the Helmholtz decomposition theorem to our network [27 , 28] . These two matrices relate to two functional components in the network architecture: The excitatory recurrence plus input , and the inhibitory recurrence . The first component provides the negative divergence that defines the gradient of computation , while the second steers its direction as follows . Since M2 has a divergence of zero , this component is rotational . If , in addition , M2 is skew-symmetric so that − M 2 = M 2 T , the system is rotating , and the eigenvalues of M2 will be imaginary only . In general , β2 ≠ −β1 and M2 is thus not skew-symmetric . However , note that a transform of the form Φf ( x ) Φ−1 can be found that makes M2 skew-symmetric . Because such transform does not change the eigenvalues of M1 , it will not change its divergence either . For above example , Φ = 1 0 0 0 1 0 0 0 β 1 β 2 ( 13 ) which will result in a version of M2 that is skew-symmetric Φ M 2 Φ − 1 = 0 0 − β 1 β 2 0 0 − β 1 β 2 β 1 β 2 β 1 β 2 0 ( 14 ) The same transformation to a different metric has to be applied to M1 as well , but as both M1 and Φ are diagonal this will leave M1 unchanged , M1 = ΦM1Φ−1 . The orderly progression through the forbidden subspaces can be understood in this framework: The negative divergence provided by M1 enforces an exponential reduction of any volume of the state space while the rotational component M2 enforces exploration of the state space by directing ( steering ) the dynamics . The stability of the permitted subspaces can also be understood in this framework . Permitted subspaces are contracting , despite the strong self-recurrence of the excitatory elements that results in positive on-diagonal elements . This high gain , which is necessary for computation , is kept under control by a fast enough rotational component M2 which ensures stability . This can be seen directly by considering one of the constraints imposed by contraction analysis on valid parameters affecting an individual neuron: keeping α < 2 β 1 β 2 guarantees that the system rotates sufficiently fast to remain stable . Note that both M1 and M2 change dynamically as a function of the currently active subspace . Thus , the direction and strength of the divergence and rotation change continuously as a function of both the currently active set as well as the input . While the network is in the expansion phase , the volume of the state space in which the dynamics of the network evolves is shrinking . This process depends on initial conditions and external inputs . Consequently , the sequence by which dimensions are removed from state space is sensitive to initial conditions . To quantify the behavior of a network for arbitrary initial conditions it is useful to compute its information entropy H ( t ) as a function of time ( see methods ) . The smaller H ( t ) , the smaller the uncertainty about which subspace the network occupies at time t . Once H ( t ) ceases to change , the network has converged . To reduce state uncertainty , and so to reduce entropy , is fundamentally what it means to compute [7] . The entropy remains 0 immediately after the application of external inputs to the network , because the same “all on” subspace is reached regardless of initial conditions . Thereafter the network begins to transition through the hierarchy of forbidden subspaces ( Fig . 2F ) . This process initially increases the entropy as the network explores different subspaces . Eventually , after attaining peak entropy , the network reduces entropy as it converges towards one of its few permitted subspaces . Fig . 3 illustrates this process for the network shown in Fig . 2A . Increasing the gain by increasing the value of the self-recurrence α increases the speed by which entropy is changed but not its asymptotic value . This means that all permitted subspaces are equally likely to be the solution but that solutions are found mode quickly with higher gain . Adding additional constraints through excitatory connections makes some permitted subspaces more likely to be the solution , and so the asymptotic entropy is lower ( see Fig . 3B , where a connection α2 = 0 . 2 from unit 1 to 2 and α3 = 0 . 2 from unit 4 to 2 was added ) . Note how the number of constraints and the gain of the network are systematically related to both the increasing and decreasing components of H ( t ) . For example , increasing the gain leads to a more rapid increase of entropy , reaching peak earlier and decaying faster towards the asymptotic value ( Fig . 3A ) . Also , adding constraints results in smaller peak entropy , indicating that the additional constraints limited the overall complexity of the computation throughout ( Fig . 3B ) . The network reaches maximal entropy when there exists the largest number of forbidden subspaces having the same divergence . This occurs always for an intermediate value of divergence , because then there occurs the largest number of subspaces having an equal number of units active and inactive . This can be seen by considering arg max k ( N k ) = N 2 ( if N is even ) , i . e . the network will have maximal entropy when the number of active units is 50% . Numerically , this can be seen by comparing the time-course of the maximal positive eigenvalue or the divergence with that of the time-course of the entropy ( Fig . 3C , D ) . Overall , H ( t ) demonstrates the dynamics of the computational process , which begins in the same state ( at its extreme , all units on ) , and then proceeds to explore forbidden subspaces in a systematic fashion by first expanding and than contracting towards a permitted subspace . We will now use the concepts introduced in the preceding sections to explain how our circuits compute and how this understanding can be utilized to systematically alter the computation through external inputs and wiring changes in the network . Provided that the external input remains constant , the network proceeds in an orderly and directed fashion through a sequence of forbidden subspaces . This sequence of steps is guaranteed to not revisit subspaces already explored , because when in a forbidden subspace S1 with divergence d1 , the next subspace S2 that the network enters must have more negative divergence d2 < d1 . It thus follows that when the system has left a subspace with divergence d1 that it can never return to any subspace with divergence ≥ d1 . It also follows that the network can only ever enter one of the many subspaces Si with equal divergence di = X ( Fig . 2F shows an example ) . Not all subspaces with lower di than the current subspace are reachable . This is because once a unit has become inactive by crossing its activation threshold , it will remain inactive . Together , this introduces a hierarchy of forbidden subspaces that the network traverses while in the exploration phase . Fig . 4A shows the hierarchy of the sets imposed by the network used in Fig . 2 . This tree-like structure of subspaces constrains computation in such a way that at any point of time , only a limited number of choices can be made . As a consequence , once the network enters a certain forbidden subspace , a subset of other forbidden and permitted subspaces becomes unreachable ( Fig . 4B ) . What those choices are depends on the input whereas the tree-like structure of the subspaces is given by the network connectivity . Knowledge of how the network transitions through the hierarchy of forbidden subspaces can be used to systematically introduce biases into the computational process . Such additional steering of the computation can be achieved by adding connections in the network . Additional off-diagonal excitatory connections , for example , will make it more likely that a certain configuration is the eventual winner . An example of this effect is shown in Fig . 5 , where adding two additional excitatory connections results in the network being more likely to arrive in a given permitted subspace than others . For identical inputs ( compare Figs . 5B and 4B ) the resulting permitted subspace can be different through such steering . The network remains continuously sensitive to changes in the external input . This is important and can be used to steer the computation without changing the structure of the network . In the absence of changes in the external input , the network is unable to make transitions other than those which lead to subspaces with lower divergence . When the input changes , on the other hand , the network can make such changes . For example , if the inputs change as a consequence of the network entering a certain forbidden subspace , the network can selectively avoid making certain transitions ( Fig . 6A ) . This will steer the computation such that some permitted subspaces are reached with higher likelihood . Noisy inputs similarly can lead to transitions which make divergence less negative . Neverthless , the large majority of transitions remains negative as long as noise levels are not too large . For example , repeating the same simulation but adding normally distributed noise with σ = 1 and μ = 0 resulted in 26% of transitions being against the gradient ( see Fig . 6B for an illustration ) . So far , we have illustrated how negative divergence and expansion jointly drive the computational process in a simple circuit of multiple excitatory neurons that have common inhibition . While this circuit alone is already capable of performing sophisticated computation , many computations require that several circuits interact with one another [23 , 29] The concepts developed in this paper can also be applied to such compound circuits , because the circuit motifs and parameter bounds we describe guarantee that collections of these circuits will also possess forbidden and permitted subspaces and are thus also computing . The compound network is guaranteed to be dynamically bounded , which means that no neuron’s activity can escape towards infinity . This property of the collective system relies on two key aspects: i ) collections of individual circuits with negative divergence also have negative divergence , and ii ) collective stability [24 , 30] . Together , these properties guarantee that collections of the motifs will compute automatically . Consider a network assembled by randomly placing instances of the two circuit motifs on a 2D plane and connecting them to each other probabilistically ( Fig . 7A , B and methods ) . This random configuration results in some excitatory elements sharing inhibition via only one inhibitory motif , whereas others take part in many inhibitory feedback loops ( Fig . 7B ) . This random circuit will compute spontaneously ( Fig . 7C , D ) . It is not known a priori how many forbidden and permitted subspaces the network has , nor how many possible solutions it can reach . Nevertheless , it is guaranteed that the network will reduce entropy and eventually reach a permitted subspace ( Fig . 7E ) . The more connections ( constraints ) that are added to the network the smaller the number of permitted subspaces , and generally the harder the computation will become . How long the computation will take to reach a permitted subspace depends on both the network size , and the number of connections ( constraints ) . Generally , the smaller the number of permitted subspaces the harder the computation will be . The important point is that random instances of such circuits will always compute , which means they will always reach a permitted subspace ( Fig . 7F ) .
The contribution of this paper has been to explore the fundamental role of instability in driving computation in networks of linear threshold units . Previous studies of computation in neural networks have focused on networks of sigmoidal units with symmetrical connectivity . Our networks of asymetrically connected LTNs draw attention to important features of computation that were not apparent in these previous models . The conditional selective behavior crucial for computation depends on the threshold nonlinearity of the LTN . However , in order to make use of these non-linearities the network must express substantial gain . Because the activation of LTNs is unbounded for positive inputs , the network can in principle produce very high activations through unstably high gain . In these networks , computation is expressed as passage through a sequence of unstable states . It is this dynamical trajectory by which the network computes [1 , 2 , 31] . Despite this essential instability , the system does not escape , but remains bounded in its behavior . In this paper we have analyzed why this is so . We find that the instabilities are self limiting , and that the overall process of computation is systematically quenched by Gaussian divergence . Contraction analysis provides explicit tools to quantify both instantaneous rates of exponential convergence to limiting states or trajectories , and divergence rates from specific subspaces . Here , we use these tools to analyze the unstable phase of the dynamics . This phase is crucial , because computation is inseparable from instability . Here we have made steps towards characterizing and explaining these phenomena . The type of dynamical system we consider can implement soft-WTA type behavior , amongst others . This makes our framework applicable to the extensive body of literature on this type of network [32–38] . While simple , the soft-WTA is a powerful computational primitive that offers the same computational power than a multi-layer perceptron [32] . Key aspects of what makes WTA-networks powerful are high network gain , which allows computations that require sparsification , and also provides stability . While the computational power of WTAs has long been recognized and exploited to implement simple cognitive behaviors [23 , 29 , 39] , it has remained unclear what it means to compute in such networks . Here , we provide such understanding in terms of a dynamical system . This system is physically realizable by realistic neurons and their connections . Other work in this direction has focused on abstract mathematic models [36 , 40–43] , and less on physically realizable dynamical computation . More recently , others [44–46] have offered useful models for understanding the principles whereby the brain may attain cognition , but these approaches do not offer methods for implementing such algorithms as physical computation in neuronal circuits . The advances of this paper can be seen in contrast with classical assumptions concerning the form of activation functions , continuous sensitivity to input , and symmetry of connections . For example , the behavior of our LTN networks can be contrasted with networks of the kind originally proposed by Hopfield [20] that allow no self-connections ( wii = 0 , ∀i ) , have symmetric connectivity ( wij = wji ) , and their activation function is bounded on both sides . This guarantees bounded dynamics by construction , allowing such networks to express high gain by virtue of a steep activation function rather than through connections of the network . However , a consequence of this is that when it operates with high gain the network operates in saturation and thus becomes insensitive to input apart from initial conditions . Such networks have neither negative divergence nor rotational dynamics , which together with insensitivity to external input severely restricts their computational abilities as well as systematic design . Importantly , our networks are continuously sensitive to their inputs . These external inputs are a combination of signal and noise and can transfer the network from one subspace to an other at any point of time and this transfer can be against the gradient imposed by negative divergence . Non-autonomous systems continuously interact with their environment , for which continuous sensitivity to input is crucial . Systems of asymmetrically interacting linear threshold units are well suited for this situation . This is because their non-saturating units make the system adaptive to the input amplitudes and sensitivity to inputs is conditional on the current state , i . e . only the inputs contributing to the dynamics of the currently active state influence the dynamics . Although there has been a considerable amount of work on symmetric networks , biological neuronal networks are always asymmetric because of inhibitory neurons . Also , the inhibitory inputs to an excitatory neuron can be substantially stronger than the excitatory inputs . This results in particularly strong asymmetry , a property with many implications for computation in such networks [47] . The theoretical study of networks with defined cell types ( excitatory or inhibitory ) thus requires asymmetric connectivity . Previous studies have used infinitely fast all-to-all inhibition to circumvent this problem , which results in symmetric connectivity but lacks defined cell types . Such networks allow dynamically bounded activity for linear threshold units [16 , 48] . Apart from being biologically unrealistic , such networks can only express limited gain and are thus computationally very limited [47] . By contrast , our networks express high gain and dynamic instabilities during the exploration phase . Their asymmetric connections provide the rotational dynamics that keep their activity bounded despite this high gain . It is worth noting that many powerful algorithms , such as e . g . the Kalman filter [49 , 50] also rely on negative feedback and strongly asymmetric connectivity . The dynamics of the exploration phase are highly structured because the different forbidden subspaces are systematically related to one another . Indeed , the subspaces are ordered in a hierarchy through which the dynamics proceed . At any point in this hierarchy only a limited and known set of subspaces can be entered next ( unless the external input changes ) . The systematic understanding of the unstable dynamics driving exploration can be used to steer and modify the computational trajectory while it is in process , rather than only when a solution has been found . The network can influence its environment continuously as a function of the forbidden subspaces it traverses , for example by executing a specific action whenever a particular subspace is entered . This feature can be used to make the computations of several networks dependent on each other . For example . to enforce dependencies between several ongoing computations such as , “all solutions must be different” . The connections of the network are the constraints imposed on the computation . The more connections per neuron , the fewer possible solutions exist and the harder ( slower ) the computation is . From this point of view , the networks we describe perform constraint satisfaction , which is a hard computational problem and which has been proposed as an abstract model of computation [51 , 52] . Connections can be inserted systematically to forward program specific algorithms and behaviors [23 , 24 , 29] , randomly or a combination thereof . Either way , the system will compute [53] , but in the former case will execute specific algorithms while in the later the algorithm is unknown . The constraints active at any point of time depend on the state of the network as expressed by the effective connectivity of the network expressed by the switching matrix . Every time the network changes state , the switching matrix changes . Dynamically , the same concept can be applied: the effective Jacobian jointly expresses all the currently activity constraints for a given state . Only if the possible state ( s ) of a network are known is it possible to determine the effective Jacobian . An important implication is that to understand the underlying algorithm that drives the computation performed by a group of neurons knowledge of the structural connectivity is not sufficient [54–56] . This is because connectivity alone does not determine the possible states of the network . The circuit motifs and parameter bounds we describe guarantee that collections of these circuits will also possess forbidden and permitted subspaces and thus are computing . By collections we mean multiple copies of the same motifs that are in addition connected to each other , as for example in the random network discussed in the results . This is important because collections of these motifs will compute automatically , a property we refer to as collective computation . This makes it practical to design large-scale computing systems without having to perform global analysis to guarantee both the type of instability required for computation as well as stability of the solutions . It is important to note that one need not commit to a particular circuit motif beyond guaranteeing that both forbidden and permitted subspaces exist in the way we define them . While a network composed of such motifs but otherwise connected randomly will always compute , the individual states do not have meaning nor is the algorithm that the network computes known . However , the states that the network proceeds through while it computes are systematically related to each other . Consequently , assigning a meaningful interpretation to a few key states will make all states meaningful . A similar approach is used in reservoir computing , where states are first created and only later assigned with meaning by learning mechanisms [57] . A key next step will be to discover how linking specific forbidden subspaces with motor actions that in turn change the input to the system allow a computation to remain continuously sensitive to the environment while it proceeds . An other next step is to discover how several interacting systems can bias each others computations systematically to reach a solution that is agreeable to all while satisfying the local constraints of each computation . The unstable positive feedback and cross inhibition that underly expansion and divergence are common motifs found in many molecular , cellular and neuronal networks [58] . Therefore all such systems follow the very simple organizational constraints that combine these motifs . This will lead such circuits to compute spontaneously and thereby to reduce their state entropy as is characteristic of living systems .
All simulations were performed with Euler integration with δ = 0 . 01 , thus τ is equivalent to 100 integration steps . All times in the figures and text refer to numerical integration steps . Unless noted otherwise , the external inputs Ii ( t ) were set to 0 at t = 0 and then to a constant non-zero value at t = 2000 . The constant value Ii was drawn randomly and i . i . d . from N ( 6 , 1 σ ) . All simulations were implemented in MATLAB . In cases where noise was added , the noise was supplied as an additional external input term N ( 0 , σ ) with σ = 1 . τ x ˙ i + G i x i = f ( I i + N ( 0 , σ ) + α 1 x i + α 2 x i − 1 + α 2 x i + 1 − β 1 x N ) ( 15 ) A new noise sample is drawn from the random variable every τ to avoid numerical integration problems . The information entropy H ( t ) is H ( t ) = − ∑ i p i ( t ) l o g 2 ( p i ( t ) ) ( 16 ) The sum is over all subspaces i and pi ( t ) is the probability of the network being in subspace i at time t , over random initial conditions ( we used 1000 runs ) . By definition , we take 0log2 ( 0 ) = 0 . The principal analytical tool used is contraction analysis [17 , 59–61] . In this section , we briefly summarize the application of contraction analysis to analyzing asymmetric dynamically bounded networks [24] . Essentially , a nonlinear time-varying dynamic system will be called contracting if arbitrary initial conditions or temporary disturbances are forgotten exponentially fast , i . e . , if trajectories of the perturbed system return to their unperturbed behavior with an exponential convergence rate . Relatively simple algebraic conditions can be given for this stability-like property to be verified , and this property is preserved through basic system combinations and aggregations . A nonlinear contracting system has the following properties [17 , 59–61] global exponential convergence and stability are guaranteed convergence rates can be explicitly computed as eigenvalues of well-defined Hermitian matrices combinations and aggregations of contracting systems are also contracting robustness to variations in dynamics can be easily quantified Consider now a general dynamical system in ℝn , x ˙ = f ( x , t ) ( 17 ) with f a smooth non-linear function . The central result of Contraction Analysis , derived in [17] in both real and complex forms , can be stated as: Theorem Denote by ∂ f ∂ x the Jacobian matrix of f with respect to x . Assume that there exists a complex square matrix Θ ( x , t ) such that the Hermitian matrix Θ ( x , t ) *T Θ ( x , t ) is uniformly positive definite , and the Hermitian part FH of the matrix F = ( Θ ̇ + Θ ∂ f ∂ x ) Θ − 1 is uniformly negative definite . Then , all system trajectories converge exponentially to a single trajectory , with convergence rate |supx , tλmax ( FH ) | > 0 . The system is said to be contracting , F is called its generalized Jacobian , and Θ ( x , t ) *T Θ ( x , t ) its contraction metric . The contraction rate is the absolute value of the largest eigenvalue ( closest to zero , although still negative ) λ = ∣λmax ( FH ) ∣ . In the linear time-invariant case , a system is globally contracting if and only if it is strictly stable , and F can be chosen as a normal Jordan form of the system , with Θ a real matrix defining the coordinate transformation to that form [17] . Alternatively , if the system is diagonalizable , F can be chosen as the diagonal form of the system , with Θ a complex matrix diagonalizing the system . In that case , FH is a diagonal matrix composed of the real parts of the eigenvalues of the original system matrix . Here , we choose Θ = Q−1 where Q is defined based on the eigendecomposition J = QΛQ−1 . The methods of Contraction Analysis were crucial for our study for the following reasons: i ) Contraction and divergence rates are exponential guarantees rather than asymptotic ( note that the more familiar Lyapunov exponents can be viewed as the average over infinite time of the instantaneous contraction rates in an identity metric ) . ii ) No energy function is required . Instead , the analysis depends on a metric Θ that can be identified for a large class of networks using the approach outlined . iii ) The analysis is applicable to non-autonomous systems with constantly changing inputs . The boundary conditions for a WTA-type network to be contracting as well as to move exponentially away from non-permitted configurations were derived in detail in [24] . They are: 1 < α < 2 β 1 β 2 1 4 < β 1 β 2 < 1 ( 18 ) | Biological systems are obviously able to process abstract information on the states of neuronal and molecular networks . However , the concepts and principles of such biological computation are poorly understood by comparison with technological computing . A key concept in models of biological computation has been the attractor of dynamical systems , and much progress has been made in describing the conditions under which attractors exist , and their stability . Instead , we show here for a broad class of asymmetrically connected networks that it is the unstable dynamics of the system that drive its computation , and we develop new analytical tools to describe the kinds of unstable dynamics that support this computation in our model . In particular we explore the conditions under which networks will exhibit unstable expansion of their dynamics , and how these can be steered and constrained so that the trajectory implements a specific computation . Importantly , the underlying computational elements of the network are not themselves stable . Instead , the overall boundedness of the system is provided by the asymmetrical coupling between excitatory and inhibitory elements commonly observed in neuronal and molecular networks . This inherent boundedness permits the network to operate with the unstably high gain necessary to continually switch its states as it searches for a solution . We propose that very simple organizational constraints can lead to spontaneous computation , and thereby to the spontaneous modification of entropy that is characteristic of living systems . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Computation in Dynamically Bounded Asymmetric Systems |
Molecular phylogenetics and phylogenomics are subject to noise from horizontal gene transfer ( HGT ) and bias from convergence in macromolecular compositions . Extensive variation in size , structure and base composition of alphaproteobacterial genomes has complicated their phylogenomics , sparking controversy over the origins and closest relatives of the SAR11 strains . SAR11 are highly abundant , cosmopolitan aquatic Alphaproteobacteria with streamlined , A+T-biased genomes . A dominant view holds that SAR11 are monophyletic and related to both Rickettsiales and the ancestor of mitochondria . Other studies dispute this , finding evidence of a polyphyletic origin of SAR11 with most strains distantly related to Rickettsiales . Although careful evolutionary modeling can reduce bias and noise in phylogenomic inference , entirely different approaches may be useful to extract robust phylogenetic signals from genomes . Here we develop simple phyloclassifiers from bioinformatically derived tRNA Class-Informative Features ( CIFs ) , features predicted to target tRNAs for specific interactions within the tRNA interaction network . Our tRNA CIF-based model robustly and accurately classifies alphaproteobacterial genomes into one of seven undisputed monophyletic orders or families , despite great variability in tRNA gene complement sizes and base compositions . Our model robustly rejects monophyly of SAR11 , classifying all but one strain as Rhizobiales with strong statistical support . Yet remarkably , conventional phylogenetic analysis of tRNAs classifies all SAR11 strains identically as Rickettsiales . We attribute this discrepancy to convergence of SAR11 and Rickettsiales tRNA base compositions . Thus , tRNA CIFs appear more robust to compositional convergence than tRNA sequences generally . Our results suggest that tRNA-CIF-based phyloclassification is robust to HGT of components of the tRNA interaction network , such as aminoacyl-tRNA synthetases . We explain why tRNAs are especially advantageous for prediction of traits governing macromolecular interactions from genomic data , and why such traits may be advantageous in the search for robust signals to address difficult problems in classification and phylogeny .
Which parts of genomes are most resistant to compositional convergence ? Which information is vertically inherited most faithfully ? Compositional stationarity and vertical ( co- ) inheritance are key , yet frequently violated , assumptions of most current approaches in molecular phylogenetics and phylogenomics [1] . Horizontal gene transfer ( HGT ) , for example , is so common and widespread that the very existence of a “Tree of Life” has been called into question [2] , [3] . Advances in understanding the history of life will require discovery of new universal , slowly-evolving phylogenetic markers that are resistant to compositional convergence and HGT . The controversial phylogeny of Ca . Pelagibacter ubique ( SAR11 ) is a case in point . SAR11 make up between a fifth and a third of the bacterial biomass in marine and freshwater ecosystems [4] . SAR11 have very small cell sizes , genome sizes , and intergenic region sizes , possibly in adaptation to extreme nutrient limitations [5] . Some recent phylogenomic studies place free-living SAR11 together in a clade with the largely endoparasitic Rickettsiales and the alphaproteobacterial ancestor of mitochondria [6] , [7] , [8] . Other studies persuasively argue that this placement is an artifact of independent convergence of SAR11 and Rickettsiales towards increased genomic A+T contents , and that SAR11 are more closely related to the free-living Alphaproteobacteria such as the Rhizobiales and Rhodobacteraceae [9] , [10] , [11] . The monophyly of SAR11 was also recently rejected [10] , [12] . Nonstationary macromolecular compositions are a known source of bias in phylogenomics [13] , . Widespread variation in macromolecular compositions may be caused by loss of DNA repair pathways in reduced genomes [15] , [11] , unveiling an inherent A+T-bias of mutation in bacteria [16] that elevates genomic A+T contents [17] , [18] . A process such as this has likely altered protein and RNA compositions genome-wide in SAR11 , and if such effects are accounted for , SAR11 appear more closely related to Rhizobiales and Rhodobacteraceae than Rickettsiales [10] , [11] . Consistent with this interpretation , SAR11 strain HTTC1062 shares , with a large clade of free-living Alphaproteobacteria that excludes the Rickettsiales , a unique and derived codivergence of features that govern recognition between tRNAHis and histidyl-tRNA synthetase ( HisRS ) [19] , [20] . This unique functionally significant synapomorphy likely arose only once in bacteria [21] and independently contradicts affiliation of SAR11 with Rickettsiales . Can the features that govern interactions between macromolecules improve phylogenomic inferences ? The two main phylogenomic “supermatrix” and “supertree” approaches [22] treat homologous sites or genes , respectively , as statistically independent data . Yet gene product interactions have known influences on their evolution . For example , amino acid substitution rates vary inversely with interaction degree ( number of interaction partners ) in proteins [23] . Furthermore , “informational” classes of genes , which mediate the expression and regulation of other genes , have more direct and indirect interaction partners on average than induced , metabolic “operational” classes of genes [24] and are less frequently exchanged across species by HGT [25] , [26] . A celebrated exception to this “complexity hypothesis” — an exception thought to prove the rule — is that of aminoacyl-tRNA synthetases ( aaRSs ) , which are “informational” housekeeping genes with high rates of HGT; this is explained because aaRSs are thought to interact primarily with only one set of tRNA isoacceptor types [27] , [28] , [29] , [30] , [31] . Although aaRSs and also tRNAs [32] can have high rates of HGT , the co-evolved features or “rules” that govern their interactions are thought to be quite resistant to lateral transfer [33] . Generally , we propose that laterally acquired gene products are more likely to adapt to new resident networks rather than to remodel those networks in accommodation of themselves . Comprehensive , accurate identification and homology mapping of features that govern macromolecular interactions remains challenging in general . tRNAs bring two distinct advantages to such an enterprise . First , the components and interactions in the tRNA interaction network are relatively highly conserved . Second and more importantly , as illustrated in Figure 1 , because all tRNAs are globally connected through general translation factors , their structures are highly conserved not only across species but also across different functional varieties of tRNAs ( “conformity” [34] ) . Each functional variety or “class” of tRNA , defined in part by which amino acid it is charged with , is distinguished by increasingly class-specific interactions with tRNA-binding proteins and other factors ( “identity” [35] ) . The uniquely contradictory requirements on tRNAs of conformity and identity makes it possible to predict the features that govern tRNA interactions by relatively simple bioinformatic analysis of genomic tRNA sequence data alone [20] . In earlier work , we developed “function logos” to predict , at the level of individual nucleotides before post-transcriptional modification , which features in tRNA gene sequences are associated to specific functional classes of tRNAs [36] . More precisely , “class” refers to a functional variety of tRNA ( such as amino acid charging or initiator identity ) . We now call our function-logo-based predictions Class-Informative Features ( CIFs ) . A tRNA CIF answers the question: “If a tRNA gene from a group of related genomes carries a specific nucleotide at a specific structural position , how informative is that feature about function , and how over-represented is that feature in a specific functional class ? ” Our estimates are corrected for biased sampling of tRNA functional classes and sample size effects [36] , and we can calculate their statistical significance [20] . In more practical terms , a tRNA CIF corresponds exactly to a single letter in the types of tRNA function logos shown in Figure 2 in the Results presented below . The “height” or fractional information of such a letter , measured in bits , is the product of conditional information of the feature about function and the normalized odds ratio of its appearance in a particular class . Thus , the greater height such a letter has , the more functionally informative that feature is , and the more it is specifically associated to a particular tRNA functional class above background expectations . We have shown that these traits , already known to have diverged across the three domains of life [37] have evolved and diverged extensively among bacteria [21] , [38] . While a single bacterial genome does not present enough tRNA sequence data to generate a statistically significant function logo , data from related genomes may be lumped together . Although this procedure assumes homogeneity , in practice features shared across taxa yield the largest signals , while phyletic variation in class-associations of features reduces signal . Function logos recover known tRNA identity elements ( i . e . features that govern specific tRNA-aaRS interactions ) [37] , [35] , and more generally , predict features governing interactions with class-specific network partners such as amidotransferases [39] . A recent molecular dynamics study on a tRNAGlu -GluRS ( Glutaminal tRNA-synthetase ) complex identified functional sites in tRNAGlu involved in allosteric signaling that couple substrate recognition to reaction catalysis in the complex [40] . The predicted sites are associated with those from proteobacterial function logos [38] . Thus , tRNA CIFs predict class-specific functional features beyond strictly tRNA identity elements alone . In this work , we show that tRNA CIFs have diverged among Alphaproteobacteria in a phylogenetically informative manner , enabling their use as signatures for classification . We validate our approach on diverse alphaproteobacterial genomes . We show that , as with other phylogenetic markers [10] , [11] , tRNAs in SAR11 and Rickettsiales have converged in base compositions , inducing an artifactual affinity between these groups when more conventional phylogenomic methods are applied to whole tRNA sequences . Our results confirm those of multiple studies that control for genomic base content variation across Alphaproteobacteria , showing that SAR11 is not a clade [10] , [12] , and that no SAR11 strains have Rickettsiales as their closest relatives [10] , [11] . Thus , tRNA CIFs are more robust to compositional convergence than the tRNA bodies in which they are embedded . Our results suggest that the best signals in genomes for deep phylogenetic problems may lie among the features that govern macromolecular interactions .
In order to characterize tRNA CIFs within Alphaproteobacteria , we reannotated alphaproteobacterial tDNA data from tRNAdb-CE 2011 [41] and pre-publication genomic data for SAR11 . For our initial studies , we set aside the SAR11 data and organized our alphaproteobacterial tDNA database taxonomically into two parts , according to whether or not source genomes contained the uniquely derived synapomorphic tRNAHis traits described previously [21] , [19] , [20] . One part corresponded to a phylogenetically coherent “RRCH clade , ” comprising the Rhodobacteraceae , Rhizobiales , Caulobacterales , and Hyphomonadaceae , which presented the derived tRNAHis traits A73 and absence of the otherwise universally conserved genetically templated G ( defined according to the so-called “Sprinzl coordinates , ” standard in the field for enumerating tRNA structural sites [42] ) . The other part corresponded to an “RSR grade” comprising the Rhodospirillales , Sphingomonadales , and Rickettsiales , which presented “normal” bacterial tRNAHis traits C73 and genomically templated G ( an “evolutionary grade” is an ancestral and paraphyletic grouping ) . Importantly , the RRCH and RSR split defined by tRNAHis traits are broadly consistent with all phylogenomic treatments of alphaproteobacterial phylogeny to date [43] , [6] , [44] , [7] , [8] , [9] , [10] , [11] . In all , we analyzed 214 alphaproteobacterial genomes presenting 11644 predicted tRNA gene sequences ( 8773 sequences unique within their respective genomes and 3064 sequences unique overall ) . Our RRCH clade data comprised 8597 tRNA genes from 147 genomes , while our RSR grade data comprised 2792 tRNA genes from 59 genomes . We analyzed 255 tRNA genes from eight SAR11 strain genomes . Seven of eight SAR11 strain genomes available to us exhibited the unique tRNAHis/HisRS codivergence traits in common with RRCH clade genomes . In contrast , strain HIMB59 presented ancestral bacterial characters in both tRNAHis and HisRS in common with the RSR grade genomes ( tRNA data not shown , HisRS data shown in Figure S1 ) . These results immediately suggested , consistent with [10] and [12] , that HIMB59 is not monophyletic with the other SAR11 strains and is affiliated with the RSR grade , while most other SAR11 strains are unrelated to the Rickettsiales and belong in the RRCH clade . In previous work , we reported the existence of fairly extensive and general divergence of tRNA Class-Informative Features ( CIFs ) between Proteobacteria and Cyanobacteria [38] . In order to investigate tRNA CIF divergence within the Alphaproteobacteria , we computed function logos [36] of the RRCH clade and RSR grade tDNA data . Qualitatively , the RRCH and RSR function logos provide visible evidence of general tRNA CIF divergence between these two groups ( comparing function logos in Figure 2 ) . To quantify these differences and exploit them to classify genomes , we formulated a quantitative measure of how well tRNAs from a given alphaproteobacterial genome match the tRNA CIFs of one group or another . Our initial simple scoring scheme sums up the differences in fractional information values or heights of features in two different function logos for two taxonomic groups if tRNAs of a given genome of the correct class carry those features ( see Figure 2 and Materials and Methods ) . To reduce bias , we used a Leave-One-Out Cross-Validation ( LOOCV ) approach , in which we recomputed the RRCH or RSR function logos for each genome to be classified by removing its own contribution to the data . In order to compare the results against those that we would get using the entire tRNA sequences , we also scored genomes using the sum of log-odds of entire sequences from tRNA-class-specific RRCH and RSR tRNA sequence profiles , also with an LOOCV approach . Typical results are shown in Figure 3 . Although the tRNA-CIF-based phyloclassifier ( Figure 3A ) was biased positively by the much larger RRCH sample size , it achieved better phylogenetic separation of genomes than the total-tRNA-sequence-based phyloclassifier based on taxon-specific tRNA profiles for different functional classes ( Figure 3B ) . The Sphingomonadales and Rhodospirillales separated in scores from the Rickettsiales in both classifiers . Most importantly , the tRNA-CIF-based phyloclassifier placed all eight SAR11 genomes closer to the RRCH clade and far away from the Rickettsiales with HIMB59 overlapping the Rhodospirillales , while the total-tRNA-sequence-based phyloclassifier placed all eight SAR11 genomes closer to the Rickettsiales . Overall , while both scoring schemes separated taxonomically distinct clades , these results show that CIFs and total tRNA data yield different signals regarding the phylogenetic placement of SAR11 genomes . Figure S2 shows the effects of different treatments of missing data in the total-tRNA-sequence-based classifier . Method “zero , ” shown in Figure 3B , is most analogous to the method used to generate Figure 3A . Method “skip” ( Figure S2B ) shows that SAR11 tRNAs share sequence characters in common with the RSR grade that are not seen in the RRCH clade . Methods “small” and “pseudo” ( Figures S2C and S2D ) show that SAR11 have sequence traits not observed in either the RSR or RRCH datasets . Divergence of tRNA CIFs between the RRCH clade and RSR grade is general and encompasses other classes besides tRNAHis . Other classes that contributed strongly to differentiated classification of RRCH and RSR genomes by the tRNA CIF-based binary classifier include tRNACys , tRNAAsp , tRNAGlu , ( symbolized “J” ) , tRNALys , and tRNATyr ( Figure 4 ) . In a manual curation of the most obvious CIF differences between RRCH and RSR , we identified traits specific to RRCH including C7-Tyr , R8-Tyr and U15∶G48-Glu , all with heights greater than 2 bits ( the height of a CIF is the height of its letter in a function logo as shown in Figure 2 , which specifically quantifies both functional information and over-representation of a CIF in tRNAs of a particular functional class and taxonomic group; please see Materials and Methods and [45] , [36] for more details ) . RSR-specific CIFs include A12-Cys and C52∶G62-Lys . These results extend the observations of [19] who discovered unusual base-pair features of tRNAGlu among members of the RRCH clade . Also , our results suggest that the unique codivergence caused by HGT of a eukaryotic-derived HisRS into an ancestor of the RRCH clade has perturbed interactions in other tRNAs , in keeping with their network coupling as shown in Figure 1 . In classes for which the RRCH and RSR groups are well-differentiated , SAR11 strain HIMB59 uniquely groups with RSR while other SAR11 strains group with RRCH , while for other tRNA classes , all putative SAR11 strains lie outside the RRCH and RSR distributions . These results imply that more diverse alphaproteobacterial genomic data are necessary to completely resolve the phylogenetic affiliation of SAR11 strains , but strongly contradict a monophyletic affiliation of SAR11 with Rickettsiales . In order to expand on this preliminary binary classification , we developed a multiway tRNA CIF-based classifier for alphaproteobacterial genomes . Instead of computing a simple difference of summed scores as before , the multiway classifier uses seven scores as its input features , in which each score sums evidence that tRNAs from a query genome match the tRNA CIFs of a specific subclade of Alphaproteobacteria . We used these summed scores to train the default multilayer perceptron ( MLP ) model implemented in WEKA [46] with ten-fold cross-validation to avoid overfitting . The MLP is the simplest nonlinear classifier able to handle the phylogenetically dependent signals in our score vectors [47] . The output of the MLP is a seven-element vector giving the classification probabilities of the query genome for each of the seven clades . Again using an LOOCV approach , each genome in our dataset classified consistently with published taxonomic positions [6] , [44] , [8] , [9] , [10] , [11] as expressed through NCBI Taxonomy , except for all eight SAR11 strains and three additional taxa recently placed in the Rhodobacteraceae based on 16S ribosomal RNA evidence: Stappia aggregata [48] , Labrenzia alexandrii [49] and the denitrifying Pseudovibrio sp . JE062 [50] ( Figure 5 ) . Our results for SAR11 are exactly consistent with those of [10]: all SAR11 strains except HIMB59 classify as Rhizobiales , while strain HIMB59 classifies as Rhodospirillales . Furthermore , Stappia , Labrenzia and Pseudovibrio classify poorly or not at all as Rhodobacteraceae . Pseudovibrio classified four times more strongly as Rhizobiales than as Rhodobacteraceae . Even excluding SAR11 , the alphaproteobacterial genomes that we analyzed vary remarkably in both tRNA gene numbers ( reflecting genome size variation ) and tRNA G+C contents . Genomic tRNA numbers vary from under 20 for highly reduced endosymbiotic genomes to over 110 , while tRNA G+C contents range from about 53% for some Rickettsiales to over 62% for Methylobacterium and Magnetospirillum ( Table S1 ) . Despite this variation , most classifications in Figure 5 were strongly and consistently statistically supported , indicating that our classifier is generally robust to base content variation of tRNAs and even deletion of entire tRNA classes . In two different bootstrap analyses , we bootstrapped sites of tRNA data in each genome to be classified , and we also filtered away small CIFs with heights bits from our models , retrained the classifier and bootstrapped sites again . Generally , the majority of bootstrap classifications matched the original dominant classifications . Alphaproteobacteria with more A+T-rich tRNAs such as members of the genus Ehrlichia classified correctly in order Rickettsiales with high probability and bootstrap values of 100 ( or an average of 92 . 5 using only CIFs with heights above 0 . 5 bits ) . At the other extreme with more G+C-rich tRNAs in the genus Methylobacteria , all strains classified correctly as Rhizobiales with a mean bootstrap value of 89 ( or 78 using only CIFs with heights above 0 . 5 bits ) . Azorhizobium caulinodans , belonging in the Rhizobiales , has G+C-rich tRNAs at 62% , and is the only representative of its genus in our study . Even in a Leave-One-Out Cross-Validation , A . caulinodans classified correctly with bootstrap values of 94 and 77 , respectively . In our CIF bootstrap analyses , SAR11 strains either had support values greater than as Rhizobiales , majority bootstrap values as Rhizobiales ( HIMB114 at with Rickettsiales at and HTCC7211 at with Rickettsiales at ) , or a plurality bootstrap value as Rhizobiales ( HIMB5 at with Rickettsiales at ) , except for HIMB59 which had a bootstrap support value of as Rhodospirillales . Full bootstrap statistics over all seven clades with these models are provided in Table S2 for SAR11 , Stappia , Labrenzia and Pseudovibrio . In a separate analysis , we also deleted each one of the 22 functional tRNA classes from the data training multiway classification ( Table S3 ) . Classification results for all of the “known” training genomes were generally highly stable to the deletion of a tRNA functional class , with a maximum of only six out of 203 genomes changing taxonomic classifications upon deletion of any one of the following tRNA functional classes: Cys , His , Arg , and Gly . When using total tRNA sequence evidence , we could not reconstruct results similar to those in Figure 5 , by either a “classical” phylogenomic supermatrix analysis of tRNAs , or using the recent novel FastUnifrac based approach specifically adapted for tRNA data [51] . In a “supermatrix” phylogenomic approach , concatenating genes for 28 isoacceptor tRNA classes from 169 species ( 2156 total sites ) and using the GTR+Gamma model in RAxML , we estimated a Maximum Likelihood tree in which all eight putative SAR11 strains branch together with Rickettsiales ( Figure S3 ) . For this analysis , in 31% of instances when isoacceptor genes were picked from a genome , we randomly picked one gene from a set of isoacceptor paralogs . However , our results did not depend on which paralog we picked . Using a distance-based approach with FastTree , we computed a consensus cladogram over 100 replicate alignments each representing different randomized picks over paralogs . As shown in a consensus cladogram ( Figure S4 ) each replicate distance tree placed all eight putative SAR11 strains together with the Rickettsiales . Widmann Et Al . ( 2010 ) [51] introduced a novel phylogenomic approach that computes a distance tree of all tRNA sequences from all genomes , and then clusters genomes using the UniFrac metric applied to that tree . Their method , although innovative , is also based on total tRNA sequence evidence . We found that it also places all SAR11 strains together with Rickettsiales ( Figure 6 ) . These results strengthen those shown in Figures 3 and S2 which suggest that tRNA CIFs exhibit a specific evolutionary signal distinct from that of tRNA sequences as a whole . Results with total tRNA sequence evidence mirror those with 16S ribosomal RNA [52] in placing all SAR11 strains together with the Rickettsiales . We suspected that it was variability in base contents of alphaproteobacterial tRNAs — caused in part by convergence of SAR11 and Rickettsiales tRNA genes to greater A+T contents — that contributed most greatly to the discrepancies in classification results between our CIF-based classifier and the phylogenomic methods using total tRNA evidence . Increases in genomic A+T in SAR11 and the Rickettsiales have driven increases in A+T content of ribosomal RNA genes [10] . We found evidence of convergence to greater A+T contents of tRNA genes as well ( Figure 7A ) . Rickettsiales and SAR11 tRNA genes are notably elevated in both A and T , and share an overall similarity in compositions distinct from those of other Alphaproteobacteria . Furthermore , a hierarchical clustering of Alphaproteobacterial families and orders based on tRNA gene base contents closely group SAR11 and Rickettsiales together ( Figure 7B ) .
We have exploited our now well-established function logo approach [36] , which predicts functional sites in tRNAs , as a means to statistically classify genomes . We have shown that our approach is more robust to tRNA base content variation than more conventional phylogenomic approaches using total tRNA evidence . While our simple scoring schemes are not interpretable as evolutionary distances , in other work we have developed evolutionary distances based on tRNA CIFs and used them to reconstruct phylogenetic trees . Our results provide strong , albeit unconventional , evidence that most SAR11 strains are affiliated with Rhizobiales , while strain HIMB59 is affiliated with Rhodospirillales . Our results are completely consistent with phylogenomic studies that control for nonstationary macromolecular compositions among Alphaproteobacteria [9] , [10] , [11] , [12] and also with a site-rate-filtered phylogenomic analysis [44] . Our CIF-based method works even though SAR11 tRNAs and Rickettsiales tRNAs have converged in base contents ( Figure 7 ) . tRNA CIFs must be at least partly robust to compositional convergence of the tRNA bodies in which they are embedded . Our results suggest that tRNA-CIF-based phyloclassification is robust to HGT of components of the tRNA interaction network . Our alphaproteobacterial phyloclassifications were highly consistent and showed no signs of misclassification of individual genomes , even though aminoacyl-tRNA synthetases ( aaRS ) are highly prone to HGT [27] , [28] , [29] , [30] , [31] including in the Alphaproteobacteria [21] , [53] , [54] . tRNAs are also known to be horizontally transferred [32] , although confident estimation of tRNA HGT rates is difficult . Even while HGT of tRNAs and tRNA-interacting proteins may be common , HGT of foreign tRNA “identity rules” governing tRNA interactions must be relatively rare . This argument is consistent with that of [33] , who argued that a horizontally transferred aaRS is more likely to functionally ameliorate to a tRNA interaction network into which it has been transferred rather than remodel that network to accommodate itself . HGT of components may also perturb a network so as to cause a distinct pattern of divergence [21] . Wang et al . [19] discuss the possibility that RRCH tRNAHis and HisRS were co-transferred into an ancestral SAR11 genome . However , this hypothesis fails to explain the correlations of many other tRNA traits of SAR11 genomes with the RRCH clade reported here . Further investigation will be needed to clarify how HGT of aaRSs and tRNAs affect the evolution of tRNA CIFs and our novel phyloclassification method . A more distant relationship between SAR11 strains and Rickettsiales actually strengthens the genome streamlining hypothesis [5] . With a placement of SAR11 within Rickettsiales , it becomes more difficult to justify how genome reduction in SAR11 occurred by a selection-driven evolutionary process rather than the drift-dominated erosion of genomes in the Rickettsiales [55] , [17] , [56] . By the same token , polyphyly of nominal SAR11 strains implies that the extensive similarity in genome structure and other traits between HIMB59 and SAR11 reported by [57] may have originated independently . Perhaps convergence in some traits is consistent with selective streamlining , which could also explain trait-sharing between SAR11 and Prochlorococcus , marine cyanobacteria also argued to have undergone streamlining [58] . The very clear signs of data limitation evident from results shown in Figures 3 , 4 , 5 and S2 imply that better taxonomic sampling will improve our results and could ultimately resolve more than two origins of SAR11-type genomes among Alphaproteobacteria . We extracted accurate and robust phylogenetic signals from tRNA gene sequences by first integrating within genomes to identify features likely to govern functional interactions with other macromolecules . Unlike small molecule interactions , macromolecular interactions are mediated by genetically determined structural and dynamic complementarities . These are intrinsically relative; a large neutral network [59] of interaction-determining features should be compatible with the same interaction network . Coevolutionary divergence — turnover—of features that mediate macromolecular interactions , while conserving network architecture , has been described in the transcriptional networks of yeast [60] , [61] and worms [62] and in post-translational modifications underlying protein-protein interactions [63] . Coevolutionary divergence of features governing tRNA interactions may be driven by ongoing recruitment of tRNA genes to new functional classes [64] . This work demonstrates that generally , divergence of interaction-governing features is phylogenetically informative . How features that govern macromolecular interactions diverge is an open question , with possibilities including compensatory nearly neutral mutations [65] , fluctuating selection [66] , adaptive reversals [67] , and functionalization of pre-existent variation [68] . Major changes to interaction interfaces may be sufficient to induce genetic isolation between related lineages , as discussed for the 16S rRNA- and 23S rRNA-based standard model of the “Tree of Life , ” in which many important and deep branches associate with large , rare macromolecular changes ( “signatures” ) in ribosome structure and function [69] , [70] , [71] . In summary , we propose that tRNA CIFs represent one of many possible different lineage-specific “shape codes” [20] among coinherited macromolecules . The concept of tRNA identity as a “second genetic code” is an old one [72] , [73] , [74] , [75] as recounted in [76] . However , by “shape code” we intend to emphasize the potentially arbitrary and co-evolveable nature of the features that underlie macromolecular interactions in specific lineages . The shape codes of macromolecular interactions within specific cellular lineages not only create a barrier to HGT of components but resist transfer even when HGT of those components occurs . Therefore , the interaction-mediating features of macromolecules may be systems biology's answer to the phylogeny problem . Perhaps no other traits of genomes are vertically inherited more consistently than those that mediate functional interactions with other macromolecules in the same lineage . In fact , the structural and dynamic basis of interaction among macromolecular components — essential to their collaborative function in a system — may define a lineage better than any of those components can themselves , either alone or in ensemble .
The 2011 release of the tRNAdb-CE database [41] was downloaded on August 24 , 2011 . From this master database , we selected Alphaproteobacteria data as specified by NCBI Taxonomy data ( downloaded September 24 , 2010 , [77] ) . Also using NCBI Taxonomy , we further tripartitioned Alphaproteobacterial tRNAdb-CE data into those from the RRCH clade , the RSR grade ( excluding SAR11 ) , and three SAR11 genomes , as documented in Supplementary data for figure 2 . Five additional SAR11 genomes ( for strains HIMB59 , HIMB5 , HIMB114 , IMCC9063 and HTCC9565 ) were obtained from J . Cameron Thrash courtesy of the lab of S . Giovannoni . We custom annotated tRNA genes in these genomes as the union of predictions from tRNAscan-SE version 1 . 3 . 1 ( with -B option , [78] ) and Aragorn version 1 . 2 . 34 [79] . We classified initiator tRNAs and tRNAIleCAU using TFAM version 1 . 4 [80] using a model previously created to do this based on identifications in [81] provided as supplementary data . We aligned tRNAs with covea version 2 . 4 . 4 [82] and the prokaryotic tRNA covariance model [78] , removed sites with more than 97% gaps with a bioperl-based utility [83] , and edited the alignment manually in Seaview 4 . 1 [84] to remove CCA tails and remove sequences with unusual secondary structures . We mapped sites to Sprinzl coordinates manually [42] and verified by spot-checks against tRNAdb [85] . We added a gap in the −1 position for all sequences and G-1 for tRNAHis in the RSR group [19] . We reannotated HisRS genes from a custom BLAST database of the eight SAR11 strain genomes using previously identified HisRS inferred protein sequences from SAR11 strains HTCC1002 , HTCC1062 and HTCC7211 and IMCC9063 downloaded from NCBI on September 27 , 2012 . Using tBLASTn from commandline BLAST version 2 . 2 . 27+ [86] , we found one match to each SAR11 strain genome , extracted these sequences and aligned them using clustalw2 ( v 2 . 0 . 11 ) [87] . Our tRNA-CIF-based binary phyloclassifier with Leave-One-Out Cross-Validation ( LOO CV ) is computed directly from function logos , estimated from tDNA alignments as described in [36] . Here , we define a feature as a nucleotide at a position in a structurally aligned tDNA , where and is the set of all Sprinzl coordinates [42] . The set of all possible features is the Cartesian product . A functional class or class of a tDNA is denoted where is the universe of functions we here consider , symbolized by IUPAC one-letter amino acid codes ( for aminoacylation classes ) , for initiator tRNAs , and for . A taxon set of genomes or just taxon set is a set of genomes , where is the set of all genomes , and is the power set of . In this work a genome is represented by the multiset of tDNA sequences it contains , denoted . The functional information of features is computed with a map from the Cartesian product of features , classes and taxon sets to non-negative real numbers . For a feature , class and taxon set , is the fraction of functional information or “height , ” measured in bits , associated to that feature , class and taxon set . This height is the product of conditional functional information of a feature ( corrected for bias due to sampling ) , times the normalized odds ratio of it appearing in a specific class [45] , see Figure S5 for more detail . In this work , for a given taxon set , a function logo is the tuple: ( 1 ) Furthermore the set of tRNA Class-Informative Features for taxon set is defined: ( 2 ) Briefly , a tRNA Class-Informative Feature is a tRNA structural feature that is informative about the functional classes it associates with , given the context of tRNA structural features that actually co-occur among a taxon set of related cells , and corrected for biased sampling of classes and finite sampling of sequences [36] . Let denote a set of Alphaproteobacterial genomes partitioned into three disjoint subsets , and with , representing genomes from the RRCH clade , the RSR grade , and the eight nominal Ca . Pelagibacter strains respectively . To execute the Leave-One-Out Cross-Validation of a tRNA CIF-based binary phyloclassifier for a genome as shown in Figure 3A , we compute a score , averaging contributions from the multiset of tDNAs in scored against two function logos and computed respectively from two disjoint taxon sets and , with . In this study , those sets are and , denoted and respectively . Each tDNA presents a set of features and has a functional class associated to it . The score is then defined: ( 3 ) As controls , we implemented four total-tDNA-sequence based binary phyloclassifiers to score a genome , shown in Figures 3B and S2 . All are slight variations in which a tRNA of class contributes a score that is a difference in log relative frequencies of the features it shares in class-specific profile models generated from and . The default “zero” scoring scheme method shown in Figure 3B is defined as: ( 4 ) where ( 5 ) is the observed frequency of feature in tDNAs of class in set , and is the frequency of tDNAs of class in set . Method “skip” corresponding to scoring scheme and Figure S2B defined as: ( 6 ) where ( 7 ) and for as before . Methods “pseudo” and “small” corresponding to scoring schemes and Figure S2C and S2D respectively: ( 8 ) where ( 9 ) where , , , for method “pseudo , ” and , for method “small , ” , where . To create Figure 7 , we computed the base composition of tRNAs aggregated by clades using bioperl-based [83] scripts , and transformed them by the centered log ratio transformation [88] with a custom script provided as supplementary data . We then computed Euclidean distances on the transformed composition data , and performed hierarchical clustering by UPGMA on those distances as implemented in the program NEIGHBOR from Phylip 3 . 6b [89] and visualized in FigTree v . 1 . 4 . For supermatrix approaches , we created concatenated tRNA alignments from 169 Alphaproteobacteria genomes ( 117 RRCH , 44 RSR , 8 PEL ) that all shared the same 28 isoacceptors with 77 sites per gene ( 2156 total sites ) . In cases where a species contained more than a single isoacceptor , one was chosen at random . Using a GTR+ model , we ran RAxML by means of The iPlant Collaborative project RAxML server ( http://www . iplantcollaborative . org , [90] ) on January 23 , 2013 with their installment of RAxML version 7 . 2 . 8-Alpha ( executable raxmlHPC-SSE3 , a sequential version of RAxML optimized for parallelization ) ( Figure S3 ) . We tested the robustness of our result to random picking of isoacceptors by creating 100 replicate concatenated alignments and running them through FastTree [91] ( Figure S4 ) . For the FastUniFrac analysis ( Figure 6 ) we used the FastUniFrac [92] web-server at http://bmf2 . colorado . edu/fastunifrac/ to accommodate our large dataset . We removed two genomes from our dataset for containing fewer than 20 tRNAs , and following [51] removed anticodon sites . Following [51] deliberately , we computed an approximate ML tree based on Jukes-Cantor distances using FastTree [91] . We then queried the FastUniFrac webserver with this tree , defining environments to be genomes of origin . We then computed a UPGMA tree based on the server's output FastUniFrac distance matrix in NEIGHBOR from Phylip 3 . 6b [89] . All tDNA data from the RSR and RRCH clades were partitioned into one of seven monophyletic clades: orders Rickettsiales ( N = 40 genomes ) , Rhodospirillales ( N = 10 ) , Sphingomonadales ( N = 9 ) , Rhizobiales ( N = 91 ) , and Caulobacterales ( N = 6 ) , or families Rhodobacteraceae ( N = 43 ) or Hyphomonadaceae ( N = 4 ) as specified by NCBI taxonomy ( downloaded September 24 , 2010 , [77] ) and documented in supplementary data for figure 7 . We withheld data from the eight nominal SAR11 strains , as well as from three genera Stappia , Pseudovibrio , and Labrenzia , based on preliminary analysis of tDNA and CIF sequence variation . Following a related strategy as with the binary classifier , we computed , for each genome , seven tRNA-CIF-based scores , one for each of the seven Alphaproteobacterial clades as represented by their function logos , using the principle of Leave-One-Out Cross-Validation ( LOO CV ) , that is , excluding data from the genome to be scored . Function logos were computed for each clade as described in [36] . For each taxon set ( with genome left out if it occurs ) , genome obtains a score defined by: ( 10 ) Each genome is then represented by a vector of seven scores , one for each taxon set modeled . These labeled vectors were then used to train a multilayer perceptron classifier in WEKA 3 . 7 . 7 ( downloaded January 24 , 2012 , [46] ) by their defaults through the command-line interface , which include a ten-fold cross-validation procedure . We bootstrap resampled sites in genomic tRNA alignment data ( 100 replicates ) and also bootstrap resampled a reduced ( and retrained ) model including only CIFs with heights greater than bits . | If gene products work well in the networks of foreign cells , their genes may transfer horizontally between unrelated genomes . What factors dictate the ability to integrate into foreign networks ? Different RNAs and proteins must interact specifically in order to function well as a system . For example , tRNA functions are determined by the interactions they have with other macromolecules . We have developed ways to predict , from genomic data alone , how tRNAs distinguish themselves to their specific interaction partners . Here , as proof of concept , we built a robust computational model from these bioinformatic predictions in seven lineages of Alphaproteobacteria . We validated our model by classifying hundreds of diverse alphaproteobacterial taxa and tested it on eight strains of SAR11 , a phylogenetically controversial group that is highly abundant in the world's oceans . We found that different strains of SAR11 are more distantly related , both to each other and to mitochondria , than widely believed . We explain conflicting results about SAR11 as an artifact of bias created by the variability in base contents of alphaproteobacterial genomes . While this bias affects tRNAs too , our classifier appears unexpectedly robust to it . More broadly , our results suggest that traits governing macromolecular interactions may be more faithfully vertically inherited than the macromolecules themselves . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"macromolecular",
"complex",
"analysis",
"sequence",
"analysis",
"systems",
"biology",
"organismal",
"evolution",
"genome",
"evolution",
"microbial",
"evolution",
"comparative",
"genomics",
"biology",
"genomics",
"evolutionary",
"biology",
"microbiology",
"b... | 2014 | tRNA Signatures Reveal a Polyphyletic Origin of SAR11 Strains among Alphaproteobacteria |
Although psychological and computational models of time estimation have postulated the existence of neural representations tuned for specific durations , empirical evidence of this notion has been lacking . Here , using a functional magnetic resonance imaging ( fMRI ) adaptation paradigm , we show that the inferior parietal lobule ( IPL ) ( corresponding to the supramarginal gyrus ) exhibited reduction in neural activity due to adaptation when a visual stimulus of the same duration was repeatedly presented . Adaptation was strongest when stimuli of identical durations were repeated , and it gradually decreased as the difference between the reference and test durations increased . This tuning property generalized across a broad range of durations , indicating the presence of general time-representation mechanisms in the IPL . Furthermore , adaptation was observed irrespective of the subject’s attention to time . Repetition of a nontemporal aspect of the stimulus ( i . e . , shape ) did not produce neural adaptation in the IPL . These results provide neural evidence for duration-tuned representations in the human brain .
Time is a fundamental property of our perception and action . Precise time-interval estimation in the range of hundreds of milliseconds is important for motor control , motion detection , and speech recognition and generation , as well as for many other complex sensory motor tasks such as playing music or dancing [1 , 2] . Previous studies of the neural representation of stimulus duration have found various forms of gradual , time-dependent changes in neural activity . For example , the temporal probability of the occurrence of upcoming events , known as the “hazard rate , ” modulates the neural firing rate in the lateral intraparietal region in monkeys [3] . Human neuroimaging studies demonstrated that such temporal modulation of firing rate during an anticipation period is reflected in a time-varying increase or decrease of the blood-oxygenation-level-dependent ( BOLD ) signal in the primary visual cortex , right supramarginal gyrus ( SMG ) , supplementary motor area ( SMA ) , right middle frontal cortex , and cerebellar vermis in humans [4 , 5] . It has also been reported that elapsed time is represented by the time-dependent ramping activity of neurons in the posterior parietal cortex in monkeys [6] . In humans , similar time-dependent increases of BOLD responses during encoding of long time intervals ( 9 and 18 s ) have been reported in the posterior insula and superior temporal cortex [7] . Although duration information is indubitably present in such time-dependent changes in neural activity , it is not known whether durations are explicitly coded in the human brain—in other words , whether there are tuning properties for specific durations . Theoretical models such as the striatal beat-frequency model [8] , the labeled-line model [9] , and the population clock model [10] have predicted that elapsed time is explicitly represented by the selective firing of a population of neurons in response to the stimulus durations to which they are tuned [11] . There is currently some behavioral evidence for the notion of such channel-based processing of time intervals [12 , 13] . However , potential neural substrates with duration-tuned response properties have not yet been documented . Here , by using functional magnetic resonance imaging ( fMRI ) adaptation , we aimed to identify brain regions that show explicit representations of stimulus duration . fMRI adaptation is based on the principle that repetition of an identical stimulus feature produces an immediate decrease in the BOLD signal by repetitive activation of the same subpopulation of neurons [14 , 15] . If a brain area contains neural populations that are sensitive to the repeated stimulus feature , the BOLD signal shows graded adaptation depending on the perceptual similarity in the stimulus feature space between consecutive presentations [16 , 17] . A large number of previous studies have shown that perception of time obeys Weber’s Law where the perceptual discriminability of two intervals depends on the ratio of their physical differences ( deviation ratio ) [18–22] , but not on the absolute differences . The deviation ratio has been used as a proxy for perceptual discriminability in adaptation paradigms elsewhere [17] . According to the population clock model that assumes explicit time representations , duration-tuned neurons are assumed to fire selectively when a time-specific neural firing pattern is submitted from another population of neurons [10] . We hypothesized that , if there exist duration-tuned neurons , they would be selectively activated by stimulus offset and show weaker BOLD responses when similar durations are repeated . We used fMRI adaptation and a broad range of stimulus durations to identify brain regions that showed adaptation to two consecutive visual stimuli of the same duration or slightly different durations . First , we identified brain areas that showed fMRI adaptation for 400 ms of stimulus duration . In the second experiment , we examined whether the brain areas that were identified in the first experiment also showed adaptation to other stimulus durations by replicating the same experiment with a slightly longer stimulus duration ( i . e . , 600 ms ) . With these two experiments , we aimed to establish generality and robustness of fMRI adaptation to specific durations . Furthermore , we asked whether such adaptation would require explicit performance of a duration estimation task or would be observed independently of the task relevance of duration estimation . For this purpose , we examined neural adaptation during both an explicit duration discrimination task and during an implicit , nontemporal shape discrimination task . Finally , in a control experiment , we confirmed that the neural adaptation that we observed in the above experiments was specific to time and was not observed in the case of repetition of a nontemporal stimulus feature ( i . e . , shape ) .
Two groups of participants performed a duration discrimination task and a shape discrimination task as instructed at the beginning of each block ( Fig 1A ) . In the tasks , participants made same-or-different judgments on two successively presented stimuli , but on different aspects of the stimuli ( i . e . , duration or shape ) . In each trial , a reference stimulus ( adaptor ) was presented for a fixed duration , followed by a test stimulus presented for a variable duration . In experiment 1 , the duration of the reference stimulus was 400 ms and that of the test stimulus was 167 , 283 , 400 , 533 , or 650 ms , whereas in experiment 2 the duration of the reference stimulus was 600 ms and that of the test stimulus was 250 , 433 , 600 , 783 , or 967 ms . The shape of the reference and test stimuli was either a square or a circle ( i . e . , the sequence presented was square-square [SS] , circle-circle [CC] , square-circle [SC] , or circle-square [CS] ) . To minimize the effect of response preparation and selection processes on the offset of the test stimulus , the key correspondences were randomized on a trial-by-trial basis and were indicated by a response cue presented after a variable interval of 1 . 0–3 . 0 s following the offset of the test stimulus . Data of 12 ( experiment 1 ) and 14 ( experiment 2 ) participants were analyzed . The proportions of “same” responses in the time task were fitted by a Gaussian function for individuals’ behavioral data . Gaussian functions were also fitted to the task performance means for presentation purpose ( Fig 1B ) . The centers of the fitted Gaussian curves did not significantly deviate from the physical durations , showing that perceived durations were not biased from physical stimulus durations in experiment 1 ( center = 399 . 3 ± 32 . 8 , t11 = −0 . 070 , p = 0 . 946; standard deviation [SD] = 130 . 0 ± 27 . 1 ) and experiment 2 ( center = 579 . 7 ± 51 . 7 , t13 = −1 . 465 , p = 0 . 167; SD = 162 . 2 ± 28 . 3 ) . To identify which brain regions showed graded adaptation to the duration of the reference stimulus ( 400 or 600 ms ) , we analyzed the offset response to a test stimulus of variable duration . We hypothesized that BOLD responses to the test stimulus would recover from adaptation according to the deviation ratio between the reference and test stimulus durations ( i . e . , [longer duration − shorter duration] / shorter duration ) ( Fig 1C ) . This analysis revealed that a number of brain areas showed the expected pattern of a neural adaptation effect , primarily in the right hemisphere ( Fig 2A and S1 Table for experiment 1 and Fig 2C and S2 Table for experiment 2 ) . Importantly , neural adaptation to both the 400-ms ( experiment 1 ) and the 600-ms ( experiment 2 ) reference duration was found in the right inferior parietal lobule ( IPL ) , corresponding to the SMG , and in the posterior temporal cortex ( PTC ) , including the middle and inferior temporal gyri ( MTG/ITG ) . Plots of the effects at the peak coordinates of the right SMG are shown in Fig 2B ( experiment 1 ) and 2D ( experiment 2 ) . Note that , however , values of the regression coefficient ( i . e . , beta ) in most of the test durations were negative , indicating that the overall right SMG activity decreased when the test stimulus had terminated and that the adaptation effect was reflected in the degree of reduction in the BOLD signal . We interpreted that this pattern of BOLD response was generated by mixed effects of general task-related suppression and duration-selective activations within a relatively small neural population . ( See Discussion for further details . ) One possible explanation for the greater response of the right IPL and PTC to “different” conditions compared with “same” conditions is that the responses reflect general same-or-different decisions and are thus not a reflection of neural adaptation to repeated stimuli of the same duration . To address this possibility , we next examined whether adaptation occurred in response to repetition of the same shape during the shape task that shared processes of same-or-different decisions with the time task . The results showed that neither the right IPL nor the PTC was sensitive to repetitions of the same shape during the shape task in either experiment 1 or experiment 2 . Plots of the beta values in the right IPL are shown in S1 Fig ( for experiments 1 [S1A Fig] and experiment 2 [S1B Fig] ) . Instead , we found a significant effect of shape repetition in the right parahippocampal gyrus , but only in experiment 1 ( see S2 Fig for further details ) . A direct comparison between duration adaptation in the time task and shape adaptation in the shape task confirmed that the right IPL and PTC adapted specifically to repetitions of the same duration , but not to repetitions of the same shape ( S3A Fig and S1 Table for experiment 1 , S3B Fig and S2 Table for experiment 2 ) . These results suggested that neural adaptation in these regions was specifically linked to repetition of a stimulus duration rather than to decision processes regarding same versus different . Although we modeled the offset responses for the test stimuli to make the design matrix comparable to that for time , shape adaptation may be observed at the onset of the test stimulus rather than offset , because unlike duration , the information about shape is already available at the stimulus onset . Thus , setting regressors at the onsets of the test stimuli might be considered more appropriate . To address this issue , we analyzed the data of experiments 1 and 2 by setting regressors at the onsets of the test stimuli instead of at their offsets . The results of this analysis were essentially the same as the original findings—no significant clusters were found for experiments 1 and 2 , but clusters in the right parahippocampal gyrus and hippocampus were significant in experiment 1 , only at slightly more liberal threshold ( p < 0 . 001 voxel-level uncorrected , cluster size > 100 voxels ) . These results further confirm our interpretation that the neural adaptation in the right IPL and PTC are specifically associated with the repetition of the identical stimulus duration . Adaptation to duration occurred not only when the participants were performing the time task but also when they were engaged in the shape task , in which duration information was not explicitly estimated . We analyzed the degree of adaptation to repetition of duration , as above , under conditions in which the participants were engaged in the shape task . In both experiment 1 and experiment 2 , neural adaptation to time during the shape task was observed in the right IPL ( including the SMG ) ( Fig 3A and 3B and S1 Table for experiment 1 , Fig 3C and 3D and S2 Table for experiment 2 ) . The clusters in the right IPL ( SMG ) overlapped with the clusters found in the explicit time estimation task ( S4A Fig for experiment 1 , S4B Fig for experiment 2 ) , indicating that neural adaptation to duration in the right IPL ( SMG ) occurred regardless of whether or not duration was explicitly estimated . One potential concern in experiments 1 and 2 was the possibility that the lack of adaptation to repetition of a shape in the right IPL and PTC could be attributed to the fact that the shapes were simply either identical or clearly different , whereas the duration variation involved finer steps across five different levels . Therefore , differences in the task design in the shape task and the level of difficulty might have contributed to our positive finding only in the time task , but not in the shape task . To make the task design and the level of difficulty for the shape task comparable to those for the time task in experiments 1 and 2 , we varied the degree of shape change in a graded manner in a new control experiment ( experiment 3 ) . The oval-shaped test stimuli were varied in width on five levels ( i . e . , very narrow [VN] , narrow [N] , same [S] , wide [W] , and very wide [VW] ) , whereas the oval-shaped reference stimulus remained the same throughout the experiment ( Fig 4A ) . In this experiment , participants performed only the shape task because the goal of this experiment was to examine whether the right IPL and PTC show adaptation to the repetition of the same shape when the stimulus shape was manipulated at a finer step of five different levels . For this purpose , we did not intermix the time task in this control experiment , as our interest was not to re-examine the effect of duration adaptation established in experiments 1 and 2 . Furthermore , we did not expect that the intermixing of the two tasks was essential for this control experiment , because the shape task was also performed in experiments 1 and 2 in separate blocks . Data from 13 participants were analyzed for this graded-shape discrimination task . The proportions of the “same” response in the shape task were fitted by a Gaussian function for individuals’ behavioral data . Gaussian functions were also fitted to task performance means for presentation purpose ( Fig 4B ) . The center of the fitted Gaussian curve showed that the perceived stimulus width was estimated as slightly narrower than the physical stimulus width ( center = 58 . 7 ± 0 . 5 , t12 = −9 . 236 , p < 0 . 001; SD = 4 . 2 ± 1 . 2 ) . One-way ANOVA showed that the task difficulty of this experiment was similar to that of the time task in experiments 1 and 2 ( F2 , 36 = 0 . 779 , p = 0 . 467 ) , suggesting that any differences in fMRI results across these experiments ( i . e . , the time task in experiments 1 and 2 and the graded-shape task in experiment 3 ) were unlikely to be due to differences in task difficulty . We expected that , if the adaptation effect found in the right IPL and PTC in experiments 1 and 2 reflected processes of general decision-making regarding same-or-different judgments , these regions would also show a graded adaptation to other stimulus features ( i . e . , the width of an oval ) ( Fig 4C ) . We found that , despite the graded change of the shape in this experiment , the right IPL ( SMG ) and PTC did not show adaptation to repetition of the same shape . As we did for experiments 1 and 2 , setting regressors at the onsets instead of the offsets of the test stimuli did not change the results . Plots of the relationship between BOLD responses and the graded shape conditions in the right SMG ( at the same coordinates as in the plots in Fig 2B and 2D ) are shown in Fig 5A and 5B . Importantly , the BOLD signal extracted from the right SMG ( Fig 5 ) showed negative regression coefficients , indicating that the right SMG showed reduction of BOLD signal due to attention focus regardless of the task ( i . e . , duration or shape discrimination task ) . These results corroborate our conclusion that neural adaptation in the right IPL ( SMG ) was specific to time and did not reflect general decision-making processes regarding same-or-different judgments . Finally , we also examined whether the repetition of the same duration during the graded-shape task showed adaptation in the right IPL . Plots of the relationship between BOLD responses and the test duration conditions at the same coordinates as in Fig 5A and 5B are shown in Fig 5C and 5D . Although the duration adaptation was not very clear at these coordinates , we instead found that the slightly more posterior part of the right IPL ( SMG ) showed a significant graded adaptation to stimulus durations ( Fig 6 ) . These results further support our findings in experiments 1 and 2 that neural adaptation to the repetition of identical duration in the right IPL ( SMG ) occurred regardless of whether the durations were explicitly estimated . Having established the specificity of neural adaptation to stimulus duration in the right IPL , we further addressed some other potential concerns . One could argue that the duration adaptation effect observed in experiments 1 and 2 might reflect neural adaptation to the fixed interstimulus interval ( ISI ) between the reference and test stimuli , but not to the duration of the reference stimulus , because the ISI ( 0 . 5 s ) was close to the duration of the “same” conditions in these experiments ( 400 and 600 ms in experiments 1 and 2 , respectively ) . To address this concern , we performed another control experiment ( experiment 4 ) with a longer ISI that varied between 1 s and 2 . 5 s ( 0 . 5-s steps ) . Another concern was that participants’ responses were biased toward “same” when a slightly shorter or longer duration was presented as the test stimulus , resulting in considerably lower performances under these conditions . To resolve this issue in experiment 4 , we also changed the number of response alternatives and the number of variations in the duration of the test stimuli . Whereas judgments of the test stimulus as “shorter” or “longer” were both classed as “different” in experiments 1 and 2 , in experiment 4 these judgments were classed individually as “shorter” or “longer” ( Fig 7A , see Materials and Methods for full details ) . The duration of the reference stimulus was either 300 ms or 450 ms; that of the test stimulus was 200 ms , 300 ms , or 450 ms for the block with a reference duration of 300 ms , and 300 ms , 450 ms , or 667 ms for the block with a reference duration of 450 ms . A response cue was presented after a variable interval of 2 . 0–3 . 5 s following the offset of the test stimuli . Data from 20 participants were analyzed in this experiment . Proportions of the “same” responses were fitted by a Gaussian function for individuals’ behavioral data ( Fig 7B ) . Statistical analyses showed that time estimation for the 300 ms block was perceived as slightly longer ( center = 325 . 5 ± 52 . 3 , t19 = 2 . 182; p = 0 . 042; SD = 99 . 6 ± 60 . 2 ) , whereas that for the 450 ms block was not biased ( center = 463 . 4 ± 43 . 6 , t19 = 1 . 380 , p = 0 . 184; SD = 134 . 3 ± 30 . 9 ) . In contrast to the time task in experiments 1 and 2 , participants showed comparable accuracy across conditions ( F1 . 88 , 35 . 77 = 0 . 610 , p = 0 . 540 ) . As in experiments 1 and 2 , we predicted that the right IPL and PTC would show adaptation to repeated stimuli of the same duration . Our analysis showed a substantial duration adaptation effect in the right IPL ( SMG ) ( Fig 8 ) , whereas the right PTC did not show such significant adaptation . The result in the right IPL is consistent with the findings in experiments 1 and 2 and supports our observation of duration-specific adaptation in the IPL . Finally , we examined the extent to which the clusters showing adaptation to repeated stimulus duration overlapped across experiments 1 , 2 , and 4 ( Fig 9 ) . We found that the clusters in the right SMG overlapped across the three experiments . In existing literature , the right SMG is often labeled as the right temporoparietal junction ( TPJ ) , which refers to the cortex at the intersection of the posterior end of the superior temporal sulcus , the IPL , and the lateral occipital cortex [23 , 24] . Since a recent TPJ parcellation study [24] suggested that the right TPJ is subdivided into three distinct regions depending on the differences in the pattern of anatomical connectivity , we examined which of the subdivisions corresponded to the overlapped area shown in Fig 9 ( white area ) . We first identified the peak coordinates of the clusters for experiments 1 , 2 , and 4 within the overlapped area ( Fig 9 ) , and then the correspondence was evaluated using a connectivity-based parcellation atlas ( http://www . rbmars . dds . nl/CBPatlases . htm ) . The result showed that all the peak coordinates within the overlapped area ( x , y , z coordinates for experiment 1: 56 , −42 , 28; experiment 2: 64 , −40 , 30; experiment 4: 54 , −42 , 28 ) were located within the anterior TPJ ( TPJa ) . These findings together strongly indicated that the right IPL ( SMG ) showed duration-specific adaptation to a broad range of stimulus durations ( i . e . , 300 , 400 , 450 , and 600 ms ) , independently of whether the task was to make same-versus-different or shorter-versus-same-versus-longer judgments .
Our results demonstrate that neural responses in the right IPL ( corresponding to the SMG ) exhibit adaptation to repeated presentations of stimuli of the same duration , regardless of whether the participants are engaged in time estimation . This finding provides empirical support for the notion that the right IPL contains neural populations tuned for particular time intervals . Dysfunction of the right SMG because of stroke [25] or virtual lesions created by transcranial magnetic stimulation ( TMS ) [26 , 27] results in impairment of time estimation , suggesting that the right SMG is crucial for estimating time intervals . Our finding of duration adaptation in the right SMG is consistent with the findings of these studies and the emerging view of the human SMG as the locus for encoding durations [26–28] . Although the representational content of time intervals in the right SMG has been unclear , our findings suggest that time intervals are represented by duration-tuned neural populations in the right SMG . The involvement of the right IPL in time perception has been reported in a number of neuroimaging studies [5 , 29–39] . Activation of the right IPL has been shown in response to stimuli of a broad range of durations ( sub- and suprasecond ) [35 , 37] and regardless of the nature of the task ( perceptual and motor timing ) [40] . It should be noted , however , that the center of typical activations in these studies was located in the dorsal part of the right IPL ( outside of the right SMG ) ( Talairach coordinates x , y , z = 40 , −44 , 38 , corresponding to 39 , −47 , 40 in Montreal Neurological Institute ( MNI ) coordinates; see meta-analysis [40] ) , whereas the right SMG we found was located more inferiorly ( MNI coordinates: [58 , −42 , 30] in experiment 1 , [62 , −34 , 32] in experiment 2 , [54 , −42 , 28] in experiment 4 ) . Moreover , in contrast to the positive BOLD response in the dorsal part of the right IPL shown in the previous studies , overall BOLD response in the right SMG in the present study was negative . Together , these results may suggest that , while the inferior part of the right IPL ( i . e . , SMG ) encodes time by duration-tuned mechanisms , the dorsal part of the right IPL may play a different role in time perception . The IPL is often labeled as TPJ , which usually refers to the cortex at the intersection of the posterior end of the superior temporal sulcus , the IPL , and the lateral occipital cortex [23 , 24] . In addition to temporal processing , the right TPJ has been implicated in various types of perception and social cognition , such as reorienting of attention [41] and attribution of mental states to others ( i . e . , a theory of mind [ToM] ) [42 , 43] . However , whether these apparently different types of cognitive functions are commonly subserved by the right TPJ or are processed in the distinct subdivisions of the right TPJ has been a matter of debate . A recent parcellation study proposed that the right TPJ is subdivided into three distinct regions based on diffusion-weighted-imaging tractography: dorsal TPJ , ventral TPJa , and ventral posterior TPJ ( TPJp ) [24] . Moreover , a resting state functional connectivity analysis showed that the TPJa and TPJp are embedded in different functional networks: the right TPJa activity interacted with the bilateral IPL , the ventral prefrontal network , and the anterior insula , which are often associated with ventral attention network [41 , 44] , while the right TPJp interacted with the posterior cingulate , the temporal pole , and the anterior medial prefrontal cortex , which are often implicated in social cognition [45] . These lines of evidence suggest that the right TPJ can be subdivided based on anatomical and functional connectivity . This study , however , also left a possibility that these two different subregions may perform similar neural computation but on different types of information ( e . g . , social versus perceptual information ) , and some neural populations located in the adjacent area might not distinguish those different types of information . This possibility seems to be supported by a recent study which demonstrated , using multivoxel pattern analysis , that spatial distance , social distance ( i . e . , familiarity ) , and temporal distance are represented by a similar activation pattern in the right TPJ [46] . The view of a "shared computational mechanism" is also compatible with the idea that the right TPJ acts as a “reorienting” system for perception and social cognition , such as attribution of another person’s mental states [23] . A number of studies have also proposed that the right TPJ is functionally homogeneous by showing that activations in the right TPJ overlap between attention reorienting and ToM tasks [42 , 47] . Future studies should address whether and how specifically the duration-tuned time representation mechanism is shared or interacts with those reorienting system and social cognition such as ToM . We found a significant overlap of duration adaptation effect in the right IPL across experiments . The overlapped area corresponded to the right TPJa that was associated with the ventral attention network in the parcellation study [24] . A large number of studies have reported that the ventral attention network is involved in stimulus-driven shift of attention . For example , the right TPJ responds to a peripheral target that appears at an unexpected spatial location [23 , 48 , 49] or to an unexpected change in stimulus features [50–53] . Based on these previous reports , one might point out that our findings of the greater BOLD responses for the deviant durations than for the repeated duration may reflect violation of prediction , as was shown in a previous face-adaptation study [54] . In other words , participants might have expected to see the test stimuli of identical duration as the reference , and the violation of that expectation by the deviant test stimulus might have produced greater activity in the right TPJ than the “same” condition . This alternative account is , however , unlikely to be the sole explanation for our results . In experiments 1 and 2 , the number of trials was the same across test durations , and the probability of occurrence for a “different” test duration was much higher than that for the “same” stimuli ( 80% versus 20% ) . Under these circumstances , it is more likely that participants expected to see a “different” duration than the “same” duration; thus , the potential effect of prediction error signals for “different” conditions should have been minimized in the present study . This is consistent with a previous study that showed effect of repetition suppression even when the “same” versus “different” response ratio was 1:3 [55] . Moreover , in our control experiment with the graded-shape task ( experiment 3 ) , we showed that the right IPL was not adapted to the repetition of the same shape , while repetition of the identical durations showed an adaptation effect in this area . The lack of a significant effect of shape adaptation in the right IPL suggests that the graded adaptation for time cannot be explained in terms of the generic concept of prediction errors . Importantly , the task design , task difficulty , number of trials , and number of subjects in experiment 3 were all comparable to the time task in experiments 1 and 2 , suggesting that the sensitivity for detecting the adaptation signal was similar across these experiments . Taken together , we suggest that our finding of duration adaptation in the right IPL is more likely to be explained by neural adaptation than by the violation of expectation ( i . e . , prediction error signal ) . This conclusion is supported by a previous neuroimaging study that showed insensitivity of the right IPL to the violation of temporal expectation , using a temporal version of the spatial cuing paradigm [56] . Training in discrimination of temporal intervals in the range of a few hundreds of milliseconds produces trained-interval-specific improvements in temporal discrimination [57] . Intriguingly , the training effect is transferred across sensory modalities [58] and from perceptual to motor timing tasks [59] in an interval-specific manner . The interval specificity of the learning effect implies that duration-selective neurons mediate the learning of time intervals . A recent fMRI study found stronger activation in the left IPL as well as the left posterior insula following intensive training in duration discrimination in the range of a few hundreds of milliseconds [60] . However , changes in right IPL activity were not reported . One possible interpretation for the seemingly discrepant results is that the training effect might be reflected in the tuning , rather than the amplitude , of activity in the right IPL . The adaptation paradigm we reported here offers a method of testing for possible changes in time-related neuronal tuning in future studies . In our recent voxel-based morphometry ( VBM ) study [61] , we reported that individual differences in the ability to discriminate stimulus durations correlated with regional gray-matter ( GM ) volume: GM volume in the bilateral anterior cerebellum was correlated with the ability to discriminate subsecond ( <1 s ) durations , while GM volume in the right IPL was correlated with suprasecond ( >1 s ) durations . The correlation between right IPL GM volumes and suprasecond discrimination thresholds was interpreted as the reflection of the ability to focus attention on continuing to track long time intervals . One might wonder why the fMRI adaptation to the repetition of subsecond durations was found in the right IPL , while correlation between regional GM volumes and discrimination thresholds in the subsecond range was absent in the VBM study [61] . To address this point , we scrutinized the data from our previous VBM study [61] . We found that , with a slightly more liberal threshold than the stringent one that was used in our VBM study [61] , the right IPL was correlated with individual differences in subsecond duration discrimination thresholds ( MNI coordinates: x , y , z = 57 , −27 , 44 , corresponding to the right SMG; the peak coordinates were identified using the Masked Contrast Images ( MASCOI ) toolbox with primary and secondary thresholds of p < 0 . 005 and p < 0 . 05 , respectively ) . This was such that a larger GM volume in this area was associated with better duration discrimination performance . This result is consistent with another recent VBM study showing that a larger GM volume in the right TPJ is associated with performance in numerical and continuous quantity ( i . e . , line length and time estimation in subsecond range ) tasks [62] , although the peak coordinates of the cluster identified in that study were found at a more posterior ventral region than ours . This additional result provides an interesting contrast in the size-performance relationship between subsecond and suprasecond time estimation: better task performance in the subsecond range was associated with larger GM volume , while better performance was associated with smaller GM volume for the suprasecond range . We speculate that this difference might reflect the different functional roles of the right IPL for subsecond and suprasecond time estimation . In our previous VBM study , we attributed the better suprasecond timing performance with smaller IPL to the better ability to focus attention on continuing to track long time intervals . On the other hand , the better performance in the subsecond time estimation associated with greater GM volume in the IPL may reflect finer duration-tuning mechanisms . The duration-specific adaptation in the right IPL suggests that individuals with greater GM volume in this region might have a greater number of neurons tuned to different durations , enhancing the ability to estimate subsecond durations . Testing this potential hypothesis would provide further insight into the anatomo-computational relationship in the right IPL . We did not find any duration adaptation in the SMA , despite its considerable involvement in various timing tasks , as shown by previous fMRI studies [40] . Importantly , a recent electrophysiological study in monkeys reported interval-tuned responses of neurons in the SMA [63] . One possible explanation for the lack of duration adaptation in the SMA in our study might be that the interval tuning found in the monkeys was specifically associated with the motor tasks employed by that particular electrophysiological study [63] . In contrast , the duration-specific adaptation in our study is likely to reflect more perceptual aspects of time estimation and is independent of the task relevance of duration estimation . Regarding the relationship between the right IPL and SMA , a recent TMS study proposed a feed-forward mechanism of temporal information from the right SMG to the SMA [27] . That study showed that TMS over the right SMG influenced the subjective perception of time intervals , as they reported earlier [26] , and that the degree of time dilation was reflected in the contingent negative variation ( CNV ) amplitude recorded from the frontocentral site in the measurement window around the stimulus offsets [27] . On the basis of an earlier report showing a positive correlation between the CNV amplitude and SMA activity as measured by BOLD signals [64] , Wiener and colleagues ( 2012 ) suggested that there was a feed-forward mechanism of temporal information from the right SMG to the SMA . Intriguingly , it has also been shown recently by electroencephalography ( EEG ) that the amplitude of the potentials evoked by the offset of comparison intervals reflects subjective time intervals better than does the amplitude of the CNV [65] . Our findings , together with these previous reports , suggest that duration information is primarily encoded in the duration-tuned neural populations in the right IPL , regardless of the task; the information is then transferred to the SMA for task-specific processing . Although a considerable number of previous studies reported the involvement of the basal ganglia and cerebellum in time perception [40 , 66 , 67] , neither of these regions were found to show consistent adaptation across different durations . The basal ganglia showed adaptation only to the repetition of 400 ms ( experiment 1 , see S1 Table ) , but this was not replicated for 600 ms in experiment 2 . These results may suggest that the duration tuning of basal ganglia is limited to the durations of around 400 ms; however , this notion requires further investigation . Although a previous theoretical study has proposed that the cerebellum may contain neurons tuned for duration [11] , a duration adaptation effect was not found in the cerebellum in our study . However , it is important to note that the cerebellum was not consistently covered by the field of view in our fMRI experiments except for experiment 4 , and therefore , the lack of positive findings in the cerebellum should be taken with caution . Although the cerebellum was covered in experiment 4 , further investigations with a broader range of reference durations would be needed to draw a firm conclusion on this issue . Our results showed negative BOLD responses at the offset of the test stimuli in the right IPL . The negative BOLD response indicates a reduction in overall neural activity [68] . While a deactivation rather than a reduction in the degree of BOLD activation may seem puzzling , similar deactivation patterns for orientation-tuned fMRI adaptation in the visual cortex have previously been reported [69] . In that study , Fang et al . investigated the effect of neural adaptation effect for orientation tuning in the visual cortex following brief adaptation versus prolonged adaptation . The study found that the neural adaptation in the visual cortex was related to suppression of the BOLD signal for repetitions with the same orientation and to a graded increase of the BOLD response as the difference in adaptor and test stimulus orientation increased . Similarly to the present study , they showed a negative BOLD response when the same orientation was repeated ( 0° ) or when the orientation of the test stimulus was only slightly different ( 7 . 5° ) from that of the adaptor . Fang et al . interpreted this negative response as “[…] maybe attributed to the overlapping neural populations tuned to 0 and 7 . 5°” [69] . The idea of the overlapped neural populations is also consistent with the computational model of duration channels [70] . We speculate that the overall negative responses observed in the present study may reflect suppression of neural activity due to attentional focus on the duration judgment task . Previous studies have reported that the right TPJ shows strong deactivations during attentionally demanding visual tasks [23 , 71–73] . Since the right TPJ has been considered to act as a “circuit breaker” that interrupts ongoing processes when a potentially task-relevant stimulus appears [41] , the deactivation of the right TPJ has been proposed to play a role in preventing attention from reorienting to task-irrelevant information when performing an attentionally demanding task [23 , 72 , 73] . Time perception is generally susceptible to interference from task-irrelevant stimulus properties ( e . g . , [34 , 70 , 74–78] ) and thus likely to require attention . The relatively low task performance in the present study also suggests that attentional demand was high during the time task in the present study . The observed negative BOLD response in the right IPL could therefore be attributed to inhibition of task-irrelevant inputs for accurate time estimation . One possible interpretation for the negative BOLD response in our study is that this deactivation stems from a mixture of neuronal populations that are deactivated during the task and neural populations that are activated for specific durations . In this interpretation , the duration-tuned neural populations are adapted by the repetition of the same duration , whereas other neurons , unrelated to time estimation , are suppressed during the task , causing the overall negative BOLD responses in the right IPL . This interpretation is in line with previous studies that attributed negative BOLD responses in the right TPJ during other types of visual tasks to the mixture of increased and decreased activities of heterogeneous neuronal populations in the right TPJ [23 , 72] . Importantly , the negative BOLD response at the offset of the test stimuli was also found in the graded-shape discrimination task in experiment 3 in which the task design and difficulty was comparable to the time task in experiments 1 and 2 . Nevertheless , variation in the degree of BOLD reduction ( i . e . , fMRI adaptation ) was uniquely found for the repetition of identical duration but not for the repetition of a shape . These results further support our interpretation that the negative BOLD response at the offset of the test duration found in the right IPL is related to the attention focus in general , while graded adaptation effect was unique to the repetition of the same duration . It has been shown that extensive adaptation to repetitions of identical durations ( e . g . , ~100 repetitions ) can produce perceptual changes in perceived duration ( “aftereffect” ) [70] . In contrast , the paired presentation paradigm with a single presentation of an adapter does not produce a perceptual change in duration as shown by a previous psychophysical study ( see Discussion in [70] ) . Although our fMRI adaptation paradigm and previous psychophysical studies of time adaptation [13 , 70] indicate the existence of duration-tuned neural representations , further investigation is required to determine whether the right IPL identified in the present study mediates the psychophysically observed aftereffect . Intriguingly , previous studies of orientation adaptation demonstrated that a prolonged presentation of an adapter produced an fMRI adaptation across multiple visual areas , whereas adaptation to a briefly presented stimulus produced adaptation only in the higher visual cortex [69 , 79] . These neuroimaging studies in orientation adaptation suggest that differences in the strength of adaptation could produce different fMRI adaptation . Direct comparison between the two different adaptation paradigms ( i . e . , single presentation versus 100 times repetition of the same duration ) would provide further insight into the relationship between duration-tuned representations in the right IPL and the psychophysical duration aftereffect . Here , we carefully ruled out potential confounding factors and replicated the main findings by using multiple experimental paradigms and four standard durations . Our findings of duration adaptation in the right IPL are attributable neither to general decision-making processes ( experiments 1 and 2 ) nor to graded manipulation of a stimulus feature ( experiment 3 ) . Moreover , we confirmed that adaptation was induced by the duration of the reference stimulus and not by the fixed interval between the reference and test stimuli ( experiment 4 ) . Differences in task difficulty between conditions or experiments were equated and thus should not have accounted for our findings of duration adaptation ( experiments 3 and 4 ) . The replication of duration adaptation in the right IPL across the experiments strongly corroborates our conclusion that this area represents the durations of salient events in a duration-selective manner . On the other hand , repetition suppression in the other brain areas such as the PTC and prefrontal cortex were not reproduced across experiments in the present study . One possible interpretation for this experiment specific effect is that such apparent repetition suppression was the result of the potential confounding factors listed above ( e . g . , general decision-making processes , graded manipulation of a stimulus feature , and fixed interval between the reference and test stimuli ) . Alternatively , neurons in those areas might specifically be tuned for a very narrow range of duration used in that particular experiment . Future studies applying single-unit recording technique for those areas ( e . g . , PTC and prefrontal cortex ) in nonhuman primates could provide clear insight into this issue . Before our work , neural correlates of time perception had been found in the form of gradual changes in neural activity over time , such as in the hazard rate of elapsed time [3 , 5] or in the form of ramping activity over time [6 , 80 , 81] . Although several theoretical models have assumed that duration is represented by populations of neurons that fire selectively for the durations to which they are tuned [8–10] , empirical evidence for such neurons in humans has until now been missing . Only a handful of electrophysiological studies in animals have reported the existence of neurons that show duration-selective responses to time intervals of a few hundreds of milliseconds; these neurons have been located in the striatum and prefrontal cortex [82] and the SMA [63] in monkeys and in the visual cortex in cats [83] . Our findings suggest that such duration-selective neurons exist in the human IPL .
All participants gave written informed consent . Experiments 1 , 2 , and 3 were approved by the University College London ( London , United Kingdom ) ethics committee , and experiment 4 was approved by the National Institute for Physiological Sciences ( Okazaki , Japan ) ethics committee . In total , 55 healthy , right-handed volunteers participated in the fMRI experiments . Seventeen volunteers participated in experiment 1 . In experiment 2 , another group of 17 volunteers was recruited . These 17 volunteers in experiment 2 also participated in experiment 3 . In experiment 4 , a new group of 21 volunteers participated . Data of five participants in experiment 1 , three participants in experiment 2 , three participants in experiment 3 , and one participant in experiment 4 were excluded from data analyses because of poor task performance in the duration or shape discrimination task ( binomial test , p > 0 . 05 ) . Moreover , one more participant was excluded from the data analysis of experiment 3 because of partial signal loss caused by a large movement of the head . Therefore , data from the remaining 12 participants ( five male and seven female , age range 19 to 28 y ) in experiment 1 , 14 participants in experiment 2 ( five male and nine female , age range 21 to 30 y ) , 13 participants in experiment 3 ( five male and eight female , age range 21 to 29 y ) , and 20 participants ( nine male and 11 female , age range 18 to 29 y ) in experiment 4 were analyzed . The proportions of the “same” responses in the time task were fitted by a Gaussian function for individuals’ behavioral data . A one-sample t-test ( α = 0 . 05 ) was performed to examine whether the center of the Gaussian was shifted from the reference durations ( experiments 1 , 2 , and 4 ) or the reference stimulus width ( experiment 3 ) . A one-way ANOVA was performed to compare the task difficulties ( i . e . , accuracy ) across experiments ( α = 0 . 05 ) . Also , a one-way repeated measures ANOVA was performed to compare the accuracy across conditions in experiment 4 ( α = 0 . 05 ) . Degree of freedom was corrected using Greenhouse-Geisser estimates of sphericity when Mauchly’s test indicated that the assumption of sphericity was violated . Visual stimuli were projected onto a half-transparent screen by an LCD projector running at 60 Hz . The screen was viewed through a mirror mounted on the head coil . Psychtoolbox ( http://psychtoolbox . org ) implemented on MATLAB software ( Mathworks , Natick , Massachusetts ) was used to present the stimuli in experiments 1 , 2 , and 3 , and Presentation Software ( Neurobehavioral System , Berkeley , California ) was used in experiment 4 . MR images for experiments 1 , 2 , and 3 were acquired with a 1 . 5-T MRI scanner ( Avanto , Siemens , Munich , Bavaria , Germany ) . Data were acquired by using a 32-channel head coil . Time-course series of 175 ( experiment 1 ) , 182 ( experiment 2 ) , and 213 ( experiment 3 ) volumes were acquired by using ascending T2*-weighted gradient-echo echo-planar imaging ( EPI ) sequences . To cover the entire cerebral cortex and basal ganglia , each volume consisted of 38 oblique slices , with 3 . 0 × 3 . 0 mm resolution , 2 . 0 mm thickness , and a 1 . 0-mm slice gap . The cerebellar cortices were not covered . The time interval between two successive acquisitions of the same slice was 3 , 230 ms , with a flip angle of 90° and a 50-ms echo time . The field of view was 192 × 192 mm . The digital in-plane resolution was 64 × 64 pixels , with a pixel dimension of 3 . 0 × 3 . 0 mm . High-resolution whole-brain MR images were also obtained by using a T1-weighted three-dimensional ( 3-D ) magnetization-prepared rapid acquisition gradient-echo ( MPRAGE ) sequence ( voxel size = 1 . 0 × 1 . 0 × 1 . 0 mm ) . The first six volumes of each fMRI run were discarded because of unsteady magnetization , and the remaining 169 , 176 , and 207 volumes per run ( a total of 676 , 704 , and 414 volumes per participant ) were used for analysis in experiment 1 , experiment 2 , and experiment 3 , respectively . MR images for experiment 4 were acquired with a 3-T MRI scanner ( Allegra , Siemens , Munich , Bavaria , Germany ) . A time-course series of 236 volumes was acquired by using ascending T2*-weighted gradient-echo EPI sequences . To cover the entire cerebral and cerebellar cortex and basal ganglia , each volume consisted of 34 oblique slices , with 3 . 0 × 3 . 0 mm resolution , 3 . 5 mm thickness , and a 0 . 56-mm slice gap . The time interval between two successive acquisitions of the same slice was 2 , 000 ms , with a flip angle of 80° and a 30-ms echo time . The field of view was 192 × 192 mm . The digital in-plane resolution was 64 × 64 pixels , with a pixel dimension of 3 . 0 × 3 . 0 mm . High-resolution whole-brain MR images were also obtained by using a T1-weighted 3-D MPRAGE sequence ( voxel size = 1 . 0 × 1 . 0 × 1 . 0 mm ) . The first five volumes of each fMRI run were discarded because of unsteady magnetization , and the remaining 231 volumes per session ( a total of 1 , 155 volumes per participant ) were used for the analysis . The fMRI data were analyzed by using statistical parametric mapping software ( SPM8; http://www . fil . ion . ucl . ac . uk/spm/ ) implemented in MATLAB . The MR images were preprocessed at the individual level . Following realignment of the fMRI data , the high-resolution 3-D T1-weighted MR images were coregistered to the fMRI data . The coregistered T1-weighted images were normalized against the MNI T1 template , and the same parameters were applied to all of the fMRI data . The anatomically normalized fMRI data were then smoothed in three dimensions by using an 8-mm full-width-at-half-maximum Gaussian kernel . Before the coregistration , slice-timing correction was applied only to the data of experiment 4 , because the TR of fMRI scan was short enough ( < 3 s ) to apply the correction . | The human brain has the ability to estimate the passage of time , which allows us to perform complex cognitive tasks such as playing music , dancing , and understanding speech . Scientists have just begun to understand which brain areas become active when we estimate time . However , it still remains a mystery how exactly the information about time is represented in the brain . In this study , we hypothesized that time might be represented by neurons that are specifically tuned to a specific duration , as has been known for simple visual features such as the orientation and the motion direction in the visual cortex . To test this idea , we performed multiple functional magnetic resonance imaging ( fMRI ) adaptation experiments in which we sought evidence of neuronal adaptation , that is , a reduction in the responsiveness of neurons to repeated presentations of similar durations . Our experiments revealed that the level of brain activity in the right inferior parietal lobule ( IPL ) was strongly reduced when a stimulus of the same duration was repeatedly presented . This finding was reproduced for a range of subsecond durations . Our results indicate that neurons in the human IPL are tuned to specific preferred durations . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Time Adaptation Shows Duration Selectivity in the Human Parietal Cortex |
While virus growth dynamics have been well-characterized in several infections , data are typically collected once the virus population becomes easily detectable . Earlier dynamics , however , remain less understood . We recently reported unusual early dynamics in an experimental system using adenovirus infection of human embryonic kidney ( 293 ) cells . Under identical experimental conditions , inoculation at low infection multiplicities resulted in either robust spread , or in limited spread that eventually stalled , with both outcomes occurring with approximately equal frequencies . The reasons underlying these observations have not been understood . Here , we present further experimental data showing that inhibition of interferon-induced antiviral states in cells results in a significant increase in the percentage of robust infections that are observed , implicating a race between virus replication and the spread of the anti-viral state as a central mechanism . Analysis of a variety of computational models , however , reveals that this alone cannot explain the simultaneous occurrence of both viral growth outcomes under identical conditions , and that additional biological mechanisms have to be invoked to explain the data . One such mechanism is the ability of the virus to overcome the antiviral state through multiple infection of cells . If this is included in the model , two outcomes of viral spread are found to be simultaneously stable , depending on initial conditions . In stochastic versions of such models , the system can go by chance to either state from identical initial conditions , with the relative frequency of the outcomes depending on the strength of the interferon-based anti-viral response , consistent with the experiments . This demonstrates considerable complexity during the early phase of the infection that can influence the ability of a virus to become successfully established . Implications for the initial dynamics of oncolytic virus spread through tumors are discussed .
The dynamics of virus spread have been studied extensively in the context of different infections , both experimentally and with mathematical models [1–3] . In particular , virus growth kinetics have been investigated in vitro and in vivo , in animal models and in human patients ( see e . g . [4–14] ) . From such data , important kinetic parameters have been measured [4 , 15–19] , such as the death rates of infected cells , the rates of viral turnover , and the basic reproductive ratio of the virus , R0 , which is thought to determine whether a successful infection can be established in a host or not . Most studies that investigate the spread of a virus through its target cell population , however , only document virus growth once the number of infected cells has already reached relatively large numbers ( in part because virus replication is hard to quantify at very low levels of infection ) . As a consequence , the dynamics during the earliest stages of virus spread remain poorly understood . Yet , this early phase can be crucial in determining the fate of the infection . We have recently studied such early dynamics experimentally in the context of adenovirus spread in vitro [20 , 21] . We tracked the spread of adenovirus infection in a 2 dimensional monolayer of human embryonic kidney ( 293 ) cells . The adenovirus used expressed green fluorescent protein , so that early virus spread from initially infected cells could be followed in space and time . A variety of interesting findings were made . Experiments showed that when virus replication initiated from a single cell , infections failed to take place for a certain fraction of the experiments . However , once at least three infected cells had been generated , a spreading infection was always established [20] . It was hypothesized that in the monolayer culture , multiply infected cells are generated relatively quickly as the number of infected cells increases , and that a high viral production from multiply infected cells could explain the lack of extinction events once three or more infected cells had been generated [20] . Following the spreading virus further ( 21 ) , two different outcomes were observed: ( i ) In what can be called a “limited spread” the infected cell population initially increased slowly , but eventually stalled at relatively low infected cell population sizes . ( ii ) In what can be called "robust spread" , the virus infection grew at a much faster pace , did not stall , and eventually reached a large number of infected cells . Importantly , these two outcomes occurred under identical experimental conditions i . e . on the same infected culture dish . In a given culture , a number of infection foci were initiated and followed , and about half of them displayed robust spread , while the other half displayed limited spread . We also note that the limited spread outcomes were not stochastic extinction events caused by random fluctuations around small numbers . Indeed , in a previous study we found that once 3 or more infected cells were generated a spreading infection was always established [20] . The occurrence of two different outcomes of early virus spread under identical experimental conditions was surprising and has remained unexplained . Here , we combine further experiments with mathematical modeling to better understand these dynamics . A relevant biological system in this respect are oncolytic viruses; they specifically replicate in cancer cells and are being explored as a treatment modality [22] . Cancers are typically infected at relatively low multiplicities , with the aim that that the virus spreads throughout the tumor cell population and kills most malignant cells . Promising results have been obtained in clinical trials , some of which involve adenoviruses upon which our experimental system is based [23] . A herpes-based virus has been approved for the treatment of melanomas , but seems to act through induction of immune responses [24] . The potential of the spreading virus itself to consistently control tumors has yet to be realized , and the initial spread from low infection multiplicities appears to be a crucial and limiting phase . Our work helps to shed light onto how this initial barrier can be overcome .
The experimental system used in this study has been described previously [20] . Human HEK 293 ( Ad-293 ) cells that express Adenovirus EIA and EIB proteins were infected in monolayer culture with a recombinant Adenovirus expressing jellyfish enhanced green fluorescent protein ( EGFP ) in place of the EIA and EIB coding region ( AdEGFPuci ) . The infections were carried out with an agar overlay , which restricted viral spread to local cell-cell spread . By conducting the infections on culture dishes with grids , it was possible to repeatedly monitor the same regions of the cultures and count the number of infected ( green ) cells over time . As described in a previous study , two types of infection outcomes were observed once an initial spreading infection was established—robust spread leading to typical virus plaques , and limited spread where the infection eventually died out ( Fig 1 ) . Under the standard infection conditions , the ratio of robust vs . limited infections was approximately 1:1 . As suggested in our previous study one possible explanation for the robust vs . limited viral infections could have been the induction of antiviral responses in the infected cells by interferon [21] . While adenovirus encodes genes that antagonize interferons [25] , this is not absolute . We therefore tested if AdEGFPuci infection of Ad-293 cells results in induction of an interferon response . When viruses infect cells , detection of viral infection by cellular sensors leads to production of interferons [25] . The interferons are released from the infected cell where they bind to interferon receptors on the surfaces of the same cell or neighboring uninfected cells; this binding leads to signal transduction and transcriptional activation of a series of IFN-responsive genes ( ISGs ) and establishment of an antiviral state . Expression of ISGs is indicative of an antiviral state . We tested RNA from AdEGFPuci-infected 293 cells for expression of several ISGs by reverse transcriptase-PCR ( RT-PCR ) . The expression of the ISG oligo-adenylate synthase 2 ( OAS2 ) was consistently induced by AdEGFPuci infection at 72 h , indicative of an antiviral state ( Fig 2A ) . In further experiments we found that the infected cells did not show induction of other ISGs tested , including OAS1 , MX-1 , IFITM1 and ISG3g , which indicated that the antiviral state induced by AdEGFPuci was partial or relatively weak . We investigated this further by studying transcriptional activation of the OAS2 promoter . 293 cells were transiently transfected with expression plasmids consisting of firefly luciferase driven by either the OAS2 promoter/enhancer , or by an artificial promoter/enhancer containing five tandem copies of the canonical interferon response element ( 5XISRE ) ( Fig 2B ) . The transfected cells were treated with interferon beta or infected with AdEGFPuci , and transcriptional activities of the reporter genes were assessed by measurement of luciferase activity at 24 and 72 hours . As expected , treatment with IFN-beta resulted in rapid induction of luciferase activity for the 5XISRE promoter by 24 h , indicative of rapid induction of an antiviral state . The effect persisted since luciferase activity was still elevated at 72 h ( compared to mock-treated cultures ) . In contrast induction of the OAS2 promoter was more modest , ca . 4-fold , compared to the > 40-fold induction for the 5XISRE promoter . In AdEGFPuci infected cells , the pattern of induction was different . The 5XISRE reporter plasmid did not show significant induction compared to the mock-treated cells at 24 h , and there was a very modest ca . 2-fold induction at 72 hr . The OAS2 reporter did not show significant increase in luciferase activity at 24 h , but there was a 5–10 fold increase at 72 hrs . Thus induction of OAS2 expression by adenovirus may involve elements in addition to the canonical ISRE . The luciferase reporter assays of Fig 2B were consistent with the RT-PCR assays of Fig 2A , and they indicate that AdEGFPuci induces a limited antiviral state in 293 cells . The fact that OAS2 expression was higher at 72 h than 24 h could reflect the time required for induction of IFN expression by the viral infection and/or modulation of the IFN response by adenoviral genes . Since the results of Fig 2 indicated that AdEGFPuci induces a limited antiviral state in 293 cells , we tested if this could be involved in the two types of viral spread observed . Valproic acid ( VPA ) , an inhibitor of histone deacetylases [26] , has been shown to inhibit the interferon response in the context of oncolytic herpesviruses [27] . As shown in Fig 2Ci , VPA reduced IFN beta induction of OAS2 RNA in Ad-293 cells at both 24 and 72 h in a dose-dependent manner . Thus it was a suitable inhibitor for these experiments . On the other hand , VPA has been shown to have both negative and positive effects on multiple cellular pathways [28] , so it could also have other effects on viral replication . Indeed , the extent of the spreading infections ( both robust and limited ) were somewhat reduced by VPA ( typical examples are shown in Figure A in the Supplementary Materials ) , and the total numbers of spreading infections were reduced in a dose-dependent fashion ( Table 1 ) . Nevertheless VPA increased the proportion of robust vs . limited spreading infections in a dose-dependent manner ( Table 1 ) . These results therefore supported the hypothesis that induction of a limited antiviral state by AdEGFPuci was influencing the relative proportion of robust vs . limited virus spread . Since VPA apparently had other effects besides inhibition of interferon on adenoviral infection , we tested two additional inhibitors of interferon: the mTOR inhibitor rapamycin [29] , and a blocking antibody to the interferon alpha receptor 2 ( IFNAR2 ) . As shown in Fig 2Cii , both anti—IFNAR2 antibody and rapamycin inhibited IFNβ induction of OAS2 RNA; inhibition by anti—IFNAR2 was rapid ( within 24 h ) , while inhibition by rapamycin was only evident at 72 hr . Experiments analogous to those of Table 1 are shown in Table 2 . Treatment with both compounds substantially increased the relative percentage of robust spreading infections , consistent with the conclusion that an interferon response was influencing the relative outcomes of the spreading infections . The effect of the anti-IFNAR2 antibody was particularly noteworthy since it would be expected to be targeting the interferon response quite specifically . In this case the shift from limited to robust spreading outcomes occurred without a change in the total number of spreading infections . Rapamycin treatment reproducibly enhanced the total numbers of spreading infections , but the mechanism of this has not been investigated . While experiments suggest that an IFN-induced antiviral state contributes to explaining our observations , we need mathematical models to test whether this is sufficient to explain the simultaneous occurrence of the limited and robust infections under identical conditions . To account for the data , a model would need to be characterized by some form of bistability , with the outcome depending on initial conditions . In a stochastic setting , the dynamics can then randomly enter one or the other domain of attraction , giving rise to different outcomes even when starting from the same initial conditions . A variety of mathematical models will be built to investigate this . Models of increasing complexity will be examined . First , we will consider ordinary differential equations ( ODEs ) that assume perfect mixing of viruses and cells . While this does not account for the spatial constraints in our experiments , it is important to start with such models for two reasons: ( i ) They are analytically more tractable , and the insights we gain from such models can be used to examine the properties of more complex , spatial models . ( ii ) Such models form the basis of much of the virus dynamics literature [1] , and this analysis will indicate whether the dynamics observed in our experiments are particular to the experimental conditions studied here , or whether this is a more broadly applicable phenomenon . Once we have analyzed such models , we will investigate the dynamics in two different , spatially explicit models: a metapopulation model that builds on the ODEs , and a 2-dimensional agent-based model that is most closely connected to our experiments . The mathematical models presented here build on a basic virus dynamics model that is well-established in the literature [1] and briefly summarized here . Denoting the number of uninfected target cells by x and the number of infected target cells by y , the model ( hereafter called model ( 1 ) ) is given by the following pair of ordinary differential equations . Because the free virus population tends to turn over fast relative to the infected cell population , free virus is assumed to be in a quasi-steady state and hence the concentration of free virus is not modeled directly . Uninfected cells are produced with a rate λ , die with a rate d , and become infected by virus with a rate β . Infected cells are assumed to die with a rate a , where a>d . This model has two equilibria: the virus extinction equilibrium where x ( 0 ) = λ/d , y ( 0 ) = 0 , and the virus persistence equilibrium where x ( 1 ) = a/β , y ( 1 ) = λ/a-d/β . In particular the virus persistence equilibrium is stable when the basic reproductive ratio of the virus ( R0 = ( λβ ) / ( da ) ) is greater than one . Such a model has been used to describe in vivo infection dynamics where target cells are produced with a constant rate . Target cell input with a constant rate , however , does not typically apply to in vitro experiments or to all in vivo tissues , where target cells can divide . Therefore , we will also consider a second version of this model that assumes division of target cells . In this version , instead of the constant input rate λ , we assume density-dependent target cell division , expressed by the term rx[1- ( x+y ) /K] , where r is the replication rate of uninfected cells and K is the carrying capacity of the system . This model will be referred to as model ( S1 ) , and more detailed properties are given in the Supplementary Materials . It is important to consider both models to investigate further whether results are limited to assumptions that apply to our experiments , or whether they have more general relevance . We extend the basic virus dynamics model ( 1 ) to include an interferon-induced anti-viral state as follows ( the modifications for model ( S1 ) are given in the Supplementary Materials ) . The uninfected and infected cell populations that are not in an anti-viral state are denoted by x1 and y1 , respectively . Uninfected cells that are in an anti-viral state are denoted by x0 . We assume that a cell that is in an anti-viral state cannot be productively infected , so infected cells in an anti-viral state are not included in this model . The model is thus given by the following set of ordinary differential equations: x˙1=λ−dx1+gx0−βx1y1−γx1y1x˙0=γx1y1−gx0−dx0y˙1=βx1y1−ay1 ( 2 ) As in model ( 1 ) , λ denotes the rate of target cell production and β the rate of infection . Uninfected and infected cells die with rates d and a , respectively . Infected cells can induce an anti-viral state in the uninfected cells with a rate γ , making them resistant to infection . This cell population is assumed to die with a rate d , and can lose its anti-viral state with a rate g . The properties of this model are very similar to those of the basic model ( 1 ) without the anti-viral state . The virus-free equilibrium is given by x1 ( 0 ) = λ/d , x0 ( 0 ) = y ( 0 ) = 0 . Virus persistence is described by the following equilibrium expressions: x1*=aβ; x0*=γ ( λβ−da ) β ( β ( g+d ) +dγ ) ; y1*=g ( λβ−da ) +βλd−d2aa ( β ( g+d ) +dγ ) . The basic reproductive ratio , R0 = ( λβ ) / ( da ) ) , is identical to model ( 1 ) . As in model ( 1 ) , there is also no bistability in this model , and when R0>1 , the virus persistence equilibrium is stable . The same properties hold for the model that assumes density-dependent division of target cells ( see model S2 in the Supplementary Materials ) . These results strongly suggest that for well-mixed systems of cells , an interferon-induced anti-viral state alone cannot explain the occurrence of two alternative outcomes under identical experimental conditions . Here , we introduce another layer of complexity into the model: Cells can be infected multiple times by the virus . Multiple infection has been shown to occur in adenovirus infections , and a higher infection multiplicity can result in higher virus output from infected cells [30–34] . If cells in an anti-viral state become infected with multiple viruses , it is possible that this can saturate factors that prevent infection of those cells . Hence , with multiple infection , cells in an antiviral state can still become infected and produce some offspring virus . Modifying our model as follows captures these features: x˙1=λ−dx1+gx0−βx1y1−γx1y1x˙0=γx1y1−gx0−βx0y1−dx0y˙1=β ( x1+y0 ) y1−ay1y˙0=βx0y1−ay0−βy0y1 ( 3 ) This model is an extension of model ( 2 ) and contains a population of infected cells that are in an anti-viral state , y0 . For simplicity we assume that two viruses within a cell are sufficient to overcome the anti-viral state . The number of viruses required to overcome the anti-viral state is unknown and can be easily adjusted in the model . Thus , when an uninfected cell in an anti-viral state , x0 , becomes infected it turns into an infected cell in an anti-viral state that fails to replicate the virus , y0 . If this cell , however , becomes infected with a second virus , it turns into a productively infected cell . Hence , multiple infection is assumed to completely overcome the anti-viral state and results in a rate of virus production that is identical to cells that are not in an anti-viral state . In the absence of the infection , we again have the trivial steady state , given by x1* = λ/d , x0* = 0 , y1* = 0 , y0* = 0 . The basic reproductive ratio of the virus is identical to that in the previous models and thus given by R0 = ( λβ ) / ( da ) . If R0<1 , then the virus-free equilibrium is stable and the only outcome is virus extinction . If R0>1 , however , the situation is more complex than in the previous models . The Supplementary Materials show that multiple non-trivial steady states are possible , which , however , are too complicated to define analytically . Hence , this is instead explored with numerical methods . We find that depending on the values of the parameters , there can be multiple non-negative stable steady states . That is , some sets of parameter values result in more than one non-negative stable steady state; while other parameter sets result in just one non-negative stable steady state . Fig 3A depicts an instance where the model is bistable: with the same set of parameters and depending on the initial conditions , the trajectories converge to different equilibria . In this figure , if the initial number of infected cells is four ( plots labeled with “1” ) the infection first takes off , but eventually stalls and regresses to a number very close to zero . If on the other hand the initial number of infected cells is five ( plots labeled with “2” ) the trajectories converge to a steady state where the overall number of cells is significantly diminished and infected cells make up a large fraction of the entire cell population . Qualitatively identical properties are found if we assume density-dependent proliferation of uninfected cells , as described in the Supplementary Materials ( see model ( S3 ) ) . A detailed numerical exploration of the parameter space for models ( 3 ) and ( S3 ) and the parameter regions that lead to multiply stable outcomes is included in the Supplementary Materials . Interestingly , in the parameter space explorations we found that parameter sets that produce bistable outcomes are more frequent when the antiviral induction rate , γ , is significantly larger than the infection rate , β ( see e . g . Figure E ( ii ) in the Supplementary Materials ) . IFN molecules are many times smaller than virions [35] , and thus should diffuse much faster than them [36] . In the context of our modeling framework this suggests significantly higher values of γ compared to β . Ultimately however , the rates β and γ should be determined experimentally in future studies . Hence , we conclude that if cells can be induced to be in an anti-viral state by IFN , and if multiple infection of cells can overcome this anti-viral state , then the ODE model can account for the two different experimentally observed outcomes . Depending on the exact initial conditions , either sustained virus growth or limited virus growth can occur for the same parameter sets . If the infection is started from the same initial number of infected cells , then it is possible that stochastic effects push the system in one domain of attraction or the other , thus leading to the alternative infection outcomes . This is explored further in the following section , using a stochastic version of our models . Here , we investigate if a stochastic formulation of the previous models can produce the different outcomes under identical initial conditions . The idea behind this hypothesis is that early stochastic events can push the system into either one of the observed infection outcomes . We implement the stochastic formulations in the standard way using Gillespie's algorithm [37] . Fig 3B depicts two stochastic simulations of model ( 3 ) , which use the same set of parameters and identical initial conditions . In the first simulation ( plot labeled with “1” ) the infection first takes off , peaks at around 60 infected cells and then starts to regress until it eventually goes extinct . In the second simulation ( labeled with “2” ) the infection is never extinguished , instead the infection persists at high levels . The same is observed for the model with density-dependent target cell proliferation ( see Supplementary Materials Figure B ) . It is important to note that the infection extinction outcomes are not in general the result of stochastic fluctuations around the initial very low numbers of infected cells . This observation is verified by Fig 3C . This figure presents the distribution of the maximum number of infected cells for simulations where the ultimate outcome was viral extinction . Note that although the initial number of infected cells is very small ( four cells ) at their maximum extension limited infections average approximately 43 cells . This behavior is even more pronounced in the simulations with the density-dependent target cell proliferation model , where the average maximum number of infected cells for the limited growth was 80 cells ( Supplementary Materials , Figure B ) . As mentioned above , inhibition of IFN signaling in our experiments resulted in a shift in outcome towards a prevalence of robust infections vs . limited infections . The same qualitative behavior is reproduced by our models . Fig 3D plots the probability of the emergence of a robust infection as a function of the relative strength of the antiviral induction rate , γ . As this figure indicates , the reduction of the strength of the antiviral induction rate , which would occur as a consequence of the down modulation of IFN signaling , results in an increased probability of establishing a robust infection . The same is observed for the model with density-dependent target cell proliferation , as described in the Supplementary Materials ( Figure B ) . To study the dynamics of multiple infection and antiviral response in a spatial setting we consider a meta-population approach , which builds closely on model ( 3 ) . The metapopulation model consists of a collection of n local patches . Within individual patches , local dynamics occur according to mass-action rules . The patches are coupled to each other by populations migrating between them . Here , we consider a one-dimensional metapopulation model where populations in a given patch can only migrate to the nearest patches . The ODE formulation of the equivalent to model ( 3 ) is given as follows: x˙1 , i=λk−dx1 , i+gx0 , i− ( β/k ) x1 , iyi , 1− ( γ/k ) x1 , iy1 , i+ ( m/2 ) ( x1 , i−1−2x1 , i+x1 , i+1 ) x˙0 , i= ( γ/k ) x1 , iy1 , i−gx0 , i− ( β/k ) x0 , iy1 , i−dx0 , i+ ( m/2 ) ( x0 , i−1−2x0 , i+x0 , i+1 ) y˙1 , i= ( β/k ) ( x1 , i+y0 , i ) y1 , i−ay1 , i+ ( m/2 ) ( y1 , i−1−2y1 , i+y1 , i+1 ) y˙0 , i= ( β/k ) ( x0 , i−y0 , i ) y1 , i−ay0 , i+ ( m/2 ) ( y0 , i−1−2y0 , i+y0 , i+1 ) ( 4 ) where k is the carrying capacity of each patch and cells with subscript i refer to cells in the ith patch . We assumed that the patches are arranged in a 1-dimensional linear array , and both target cells and infected cells can migrate to the neighboring patches to the left and to the right of a given patch with migration rate equal to m ( the boundary conditions at i = 0 and i = n prevent cell migration outside of the array ) . The stochastic implementation of the metapopulation model follows directly from the reformulation of model ( 4 ) using Gillespie's method . We find that in the stochastic metapopulation model the two types of infection outcomes ( limited vs robust infection ) can occur under identical initial conditions and with the same set of parameters ( Fig 4B and 4C ) . Fig 4A illustrates the spatial spread of a robust viral infection . The effect of inhibiting IFN signaling ( as done in the experiments ) is represented in model ( 4 ) through a reduction of the anti-viral induction rate ( γ ) . In agreement with experiments , in the spatial metapopulation model the down modulation of IFN signaling results in an increased proportion of the robust infection outcomes ( Fig 4D ) . The same qualitative results can be obtained in a spatial metapopulation model if we assume density-dependent target cell proliferation ( see Supplementary Materials , Figure C ) . These explorations show that the induction of an anti-viral state and the ability of multiple infection to overcome the anti-viral state can account for the two infection outcomes observed in the data in a spatial setting , which is closer to our experimental conditions . The advantage of the metapopulation model is that we can build on the insights obtained from the analysis of the ODEs and explore this in a spatial setting . In the next section , we take this work further away from the ODE approach and consider a 2-dimensional stochastic spatial agent-based model of these dynamics . In the experiments , cells are arranged in a two-dimensional layer , and an agar overlay prevents long-range spread of the virus away from infected cells in the culture medium , creating conditions where virus spread is most likely to occur to neighboring cells . To model these dynamics we create a two-dimensional stochastic agent-based model , where each cell occupies a certain position in a two-dimensional rectangular lattice . There are four types of cells: susceptible uninfected cells ( x1 ) , uninfected cells in antiviral state ( x0 ) , infectious cells ( y1 ) and infected cells in antiviral state ( y0 ) . The basic rules are the same as those in the previously explored model . Because this is a model that directly relates to our adenovirus experiments , we will only consider the assumption of density-dependent target cell growth . The rules are described as follows: As we previously mentioned , in the experiments viral spread is restricted to cell neighbors . In the model the neighborhood of a cell is determined by the lattice coordinates that the cell occupies . For the simulations we choose a neighbor radius of two cells in any direction . The model is implemented using the Next Reaction Method [43] ( for details see Supplementary Materials ) . Consistent with the other models , we observed both limited and robust infection outcomes ( depicted in Fig 4E–4G ) , and that both can occur stochastically under identical parameter combinations and initial conditions . This again required the occurrence of an antiviral state , and the assumption that multiple infection can overcome the antiviral state . The effect of inhibiting IFN signaling is modeled through a reduction of the anti-viral induction rate γ . In agreement with experiments , the down modulation of IFN signaling results in an increased proportion of the robust infection outcomes ( Figure D in Supplementary Materials ) . Statistical results from the simulations and a numerical exploration of the parameter space are found in the Supplementary Materials .
Typically , when virus growth dynamics are investigated , even the early exponential growth phase corresponds to infected cell numbers that are relatively high , where virus levels are readily detectable . This applies to in vitro experiments and even more so to in vivo studies . Our work , however , has shown that complex dynamics can occur when virus concentrations are much lower ( starting from a single infected cell ) , and that these dynamics might play an important role in determining the fate of the infection . In this respect , our experiments identified IFN-induced antiviral states in cells as a crucial mechanism that contributes to the observations . Our mathematical models , however , suggest that this alone cannot explain the experimental outcomes and that additional mechanisms need to be invoked to account for the data . In this respect , we identified the ability of the virus to overcome the anti-viral state by multiple infection as a possible candidate mechanism . Under these assumptions , an initial race between the spread of the virus population and the spread of the antiviral state can stochastically result in two different outcomes under identical conditions . While this is a likely mechanism , and while there is indication that multiple infection of cells might indeed play an important role in adenovirus spread [30–34] , we have not been able to explicitly test this notion so far . Hence , one has to be aware that there could potentially be other mechanisms that might also explain the data . An important question concerns the generality of these notions , i . e . whether the results have relevance beyond the virus-cell system considered in our study . Our mathematical models suggest that such dynamics could be a more general phenomenon . Our experimental system involved a two-dimensional monolayer of cells with agar layover , which is best described by a spatially restricted agent based model . Qualitatively identical results , however , are seen in relatively simple ordinary differential equation models that describe virus dynamics in a setting without any spatial restrictions . Further , the results remain robust in different model formulations . For example , the exact target cell dynamics ( production vs . division of target cells ) does not appear to change the notions reported here . From an empirical point of view , it remains to be investigated whether multiple infection can saturate an antiviral state in cells , and how general a phenomenon this is . This mechanism can explain the data in the context of our model , and is therefore a model-generated hypothesis . In our experimental system , we demonstrated that AdEGFPuci induced only a limited antiviral state in 293 cells . It is possible that a limited anti-viral state can be overcome more easily by multiple infection than a stronger anti-viral state that may occur in other virus-host cell systems . It would be interesting to test this notion experimentally . Beyond improving our understanding of the principles of virus dynamics , our results have important practical implications for the field of oncolytic virus therapy of tumors [22 , 23 , 44] , and add to previous mathematical modeling work that analyzes the dynamics of oncolytic viruses , e . g . [21 , 45–49] . Some clinically important oncolytic viruses are based on the adenovirus used in our experiments [23] . An important first step in successful oncolytic virus therapy is that the virus establishes a robust infection and efficiently spreads throughout the tumor . The very early spread dynamics might be a crucial phase in this respect , and might pose the first barrier to success . Our study indicates that not only the strength of interferon-induced anti-viral states might be important in this respect , but that a high local infection multiplicity at the earliest stages of the infection process might be equally important for ensuring that this early barrier is crossed and that the infection enters a regime in which robust growth to large viral population sizes is achieved . As a next step , it would be useful to test these notions in the context of different , specific oncolytic viruses that grow on tumor cell lines in vitro . As mentioned above , it is possible that our findings are the result of a limited IFN-induced anti-viral response in 293 cells infected with our virus AdEGFPuci . Hence , it would be interesting to compare the oncolytic virus dynamics in the context of virus-tumor systems in which the strength of an IFN-induced antiviral state varies . For example , some tumors show abnormalities and defects in IFN signaling , and this can in fact be a mechanism for the selective replication of oncolytic viruses in tumor cells [50] . Oncolytic vesicular stomatitis virus ( VSV ) is an example of this [51] . It would be important to investigate the dynamics studied here in the context of this virus , as well as with other oncolytic viruses that are characterized by different mechanisms of tumor selectivity , and that experience the induction of a stronger IFN-induced antiviral state in tumor cells . Such insights could guide future work that aims to optimize the virus itself as well as the method by which the virus is delivered to the cancer .
Ad-293 cells , derivatives of HEK293 cells that express adenovirus EIA and EIB proteins , were purchased from Agilent Technologies ( La Jolla , CA ) as described previously [20] . AdEGFPuci , a recombinant adenovirus expressing enhanced jellyfish green fluorescent protein in place of EIA and EIB was also described previously [20] . Stocks of 109−1011 pfu/ml were used . Infections of AdEGFPuci onto Ad-293 cells on gridded culture dishes were performed at different multiplicities under conditions of plaque formation as described previously [20] . Infections were monitored by fluorescence microscopy for GFP; areas of infection were counted and scored as limited or robust . The same dishes were scored at different days post-infection . Total RNA from infected or uninfected Ad-293 cells was extracted using TRIzol ( Life Technologies ) , and 2 μg of RNA was digested with DNAse I and converted to cDNA using the qScript cDNA synthesis kit ( Quantas ) according to the manufacturer’s instructions . For detecting induction of cellular interferon response genes , PCR primer sets for different cellular genes from the Interferon Response Detection Kit ( Systems Biosciences ) were employed and PCR amplifications for different cycles were carried out according the manufacturer’s instructions . PCR products were visualized by agarose gel electrophoresis and ethidium bromide staining . For quantitative RT-PCR of OAS2 RNA , the following human OAS2 primers were used: 5’AGCTCCTCCTTTTTCCTTCCAGTC3’ ( forward ) and 5’TGGCTGGCTGCTGGCATAGAG3’ ( reverse ) . For standardization , the following human GAPDH primers were used: 5’CAACTACATGGTCTACATGTTC3’ ( forward ) and 5’ctcgctcctggaagatg3’ ( reverse ) . Quantitative RT-PCRs were performed using Power SYBR green PCR master mix with the 7900HT Fast real-time PCR system ( Applied Biosystems ) according to the manufacturer’s instructions . All qRT-PCRs were run in triplicate . The RNA expression levels were determined by the relative comparative threshold cycle ( Cr ) method . To measure transcriptional activities , the following firefly luciferase reporter plasmids were used . Path Detect ISRE-luc ( Agilent technologies ) is a luciferase reporter driven by 5 tandem copies of an Interferon response element ( ISRE ) , and is referred to here as 5XISRE-luc . The matched plasmid lacking promoter or enhancers , pCIS-CK ( Agilent ) , was used as a negative control . A luciferase reporter driven by the upstream control elements of the 2’-5’- oligo ( A ) synthetase 2 ( OAS2 ) gene was generated by first PCR amplifying OAS2 sequences ( –880 to +197 ) from 293 cell DNA using synthetic primers containing sites for Hind III and Bgl II at either end . The PCR fragment was digested with Hind III and Bgl II , and cloned into the pLuc-MCS plasmid ( Agilent ) between Hind III and Bgl II sites in the multiple cloning site of the plasmid . This plasmid was designated OAS2-luc . Ad-293 cells were transfected with the different luciferase reporter plasmids as described previously [52] , with or without interferon treatment or infection with AdEGFPuci . Cell lysates were prepared and analyzed for luciferase activity as described previously [52] , using the dual luciferase assay kit ( Promega ) . Luciferase activities were read in a Sirius-L luminometer ( Berthold Detection Systems ) ; transfections were performed in triplicate , and each assay was performed at least three times . Human interferon beta was purchased from Pepro Tech , Interferon alpha receptor 2 antibody from Life Techologies , valproic acid from Sigma-Aldrich , and rapamycin from Fisher Scientific . | We investigate in vitro adenovirus spread starting from the lowest infection multiplicities . This phase of virus dynamics remains poorly understood and is likely critical for ensuring that engineered oncolytic viruses successfully spread and destroy tumors . We find unexpectedly complex dynamics , which are analyzed with a combination of experiments and mathematical models . The experiments indicate that the induction of an interferon-based anti-viral state is a crucial underlying mechanism . The mathematical models demonstrate that this mechanism alone cannot explain the experiments , and that additional mechanisms must be invoked to account for the data . The models suggest that the ability of the virus to overcome the anti-viral state through multiple infection of cells might be one such mechanism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"and",
"Conclusion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"luciferase",
"antiviral",
"immune",
"response",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"pathogens",
"cancer",
"treatment",
"immunology",
"population",
"dynamics",
"microbiology",
"enzymology",
"oncolytic",
"virus... | 2017 | Complex Dynamics of Virus Spread from Low Infection Multiplicities: Implications for the Spread of Oncolytic Viruses |
Most bacteria live in colonies , where they often express different cell types . The ecological significance of these cell types and their evolutionary origin are often unknown . Here , we study the evolution of cell differentiation in the context of surface colonization . We particularly focus on the evolution of a ‘sticky’ cell type that is required for surface attachment , but is costly to express . The sticky cells not only facilitate their own attachment , but also that of non-sticky cells . Using individual-based simulations , we show that surface colonization rapidly evolves and in most cases leads to phenotypic heterogeneity , in which sticky and non-sticky cells occur side by side on the surface . In the presence of regulation , cell differentiation leads to a remarkable set of bacterial life cycles , in which cells alternate between living in the liquid and living on the surface . The dominant life stage is formed by the surface-attached colony that shows many complex features: colonies reproduce via fission and by producing migratory propagules; cells inside the colony divide labour; and colonies can produce filaments to facilitate expansion . Overall , our model illustrates how the evolution of an adhesive cell type goes hand in hand with the evolution of complex bacterial life cycles .
As illustrated by the examples above , cells can colonize a wide range of surfaces: including the air-liquid interface [26 , 41 , 42] , air-solid interface and liquid-solid interface–e . g . plant roots [43–46] , soil particles [47 , 48] , fungi [49] . At these surfaces , attachment is governed by distinct biophysical mechanisms , although generally speaking adhesive cells are acquired . For simplicity , our model ignores the biophysical details of attachment and simply assumes that adhesive cells can adhere to the surface , whereas non-adhesive cells cannot . As such , the model does not resemble any specific type of surface colonization , instead we aim to make a first step in exploring how adhesive cell types evolve in a dynamical environment where cells can attach and detach from a surface at any moment in time . We assume that the model consists of two environments: the liquid and the surface ( Fig 1 ) . At the onset of evolution , cells only occur in the liquid . Cells can express two cell types: sticky and non-sticky cells . Sticky cells have a reduced cell division rate ( R ) and are required for surface attachment . They can attach to any unoccupied position on the surface . Non-sticky cells can also attach to the surface , but only when immediately neighbouring a sticky cell . In other words , non-sticky cells can hitchhike with sticky cells , like observed in the lab ( e . g . S1 Text , S1C and S1D Fig ) . The surface consists of a two-dimensional hexagonal grid , so each sticky cell can have at most six non-sticky neighbours ( Fig 1 ) . Surface attachment is beneficial , because it allows cells to escape from competition in the liquid . This benefit is present as long as there is space available on the surface . At the same time , surface attachment requires sticky cells that carry the cost of a lower cell division rate . Non-sticky cells that attach to the surface by hitchhiking with sticky cells escape from the costs of being sticky , but still have the benefits of surface attachment . We examine the evolution of sticky cells for three model variants . These variants differ with respect to the differentiation strategy that evolves ( Fig 1 ) : cells either have a ( 1 ) pure strategy , ( 2 ) probabilistic strategy , ( 3 ) decision-making strategy . In the pure strategy , cells can only switch between being sticky and non-sticky by mutations . As a result , each genotype expresses one phenotype . In the probabilistic strategy , cells differentiate with a certain probability ( P ) . This probability can change over evolutionary time , by the accumulation of mutations ( for details see Material and Methods ) . In the decision-making strategy , cells can differentiate in response to the environment . Cells sense two environmental cues: the niche in which they occur ( N = 0 in the liquid and N = 1 on the surface ) and the fraction of sticky cells ( i . e . stickiness , S ) . On the surface , cells only sense the fraction of sticky cells in the neighbouring positions on the grid . In the liquid , the fraction of sticky cells is determined with respect to the entire population . The sensory input to a cell is weighted by so-called connection weights ( W1 and W2 ) . When the sum of regulatory input exceeds a given threshold ( θ ) a cell differentiates to a sticky cell . Over evolutionary time the connection weights and activation threshold can evolve ( see Fig 1 and Material and Methods ) . For each model variant , we start evolution with a population of non-sticky cells in the liquid . All genotypic variables are set to zero ( model variant 2: P = 0 , and model variant 3: W1 = W2 = θ = 0 ) . Cells can evolve for 400 . 000 time step . At each time step , one of the following events can occur ( see Material and Methods ) : ( i ) migration to the surface , ( ii ) migration to the liquid , ( iii ) cell differentiation , ( iv ) cell death , ( v ) cell division . The event that occurs is chosen randomly . We explore the outcome of evolution by varying two modelling parameters: R and Pm . R is the relative cell division rate of sticky cells . When the costs of being sticky are high , sticky cells cannot divide ( R = 0 ) and , when the costs are low , sticky and non-sticky cells are equally likely to divide ( R = 1 ) . Pm is the probability to migrate to the surface . As default setting Pm = 0 . 1 , which means that cells have a 10% probability to migrate to a random location on the surface . This does not mean that they necessarily attach to this particular location . A cell can only attach when the randomly chosen position on the surface is vacant and the cell is sticky or surrounded by a sticky cell .
We first examined the evolution of surface colonization at various relative growth rates of the sticky cells ( R = 0 , 0 . 4 , 0 . 8 , 1 ) . Fig 2 shows some representative surfaces at the end of evolution , for the pure , probabilistic and decision-making strategy . Sticky cells are shown in red and the non-sticky cells in blue . The three differentiation strategies differ in their capacity to colonize the surface ( see also Fig 3A ) . The pure strategy shows some surface colonization at all cost levels , but the number of cells on the surface is very low at high costs of being sticky , i . e . low cell division rates of sticky cells ( R ) . The few sticky cells that occupy the surface at R = 0 express a maladaptive phenotype , because these cells cannot reproduce , nor can they switch phenotype ( in the pure strategy , cells can only switch phenotype through mutations that occur during cell division ) . The probabilistic strategy evolved a nearly full surface colonization at most costs , but cannot colonize the surface at the highest costs ( R = 0 ) . The decision-making strategy can colonize the surface at all costs; even when sticky cells cannot divide ( R = 0 ) . The fraction of sticky cells on the surface decreases for higher costs of being sticky . Moreover , the fraction of sticky cells is lower in colonies from the decision-making strategy than in colonies from either the probabilistic or pure strategy . In the decision-making strategy , cells furthermore show spatial pattern formation ( Fig 2 ) . At R = 0 and R = 0 . 4 , sticky cells only have non-sticky neighbours . In other words , the sticky cells–together with their non-sticky neighbours–form separated islands on the surface . At R = 0 . 8 , sticky cells also form filaments . Filaments are short concatenations of sticky cells ( 2–8 sticky cells ) , which are surrounded by non-sticky neighbours . Only in a few cases do sticky cells also clump together . When there are no costs of being sticky ( R = 1 ) , all cells express the sticky phenotype and there is no spatial pattern formation . There are no regular spatial patterns for the probabilistic and pure strategy , because these strategies cannot account for the number of sticky neighbours . Next , we examined the fraction of sticky cells on the surface and in the liquid ( Fig 3B ) . On the surface , as shown by Fig 2 , the decision-making strategy produces the lowest fraction of sticky cells . In the pure and probabilistic strategies , the fraction of sticky cells gradually increases with higher R values . In other words , at lower costs of being sticky , a larger fraction of cells expresses the sticky phenotype . In the decision making strategy , the fraction of sticky cells does not change gradually with R . Instead , the fraction of sticky cells is around 20% ( R = 0–0 . 72 ) , 30% ( R = 0 . 8–0 . 96 ) or 100% ( R = 1 ) . These levels correspond to distinct spatial patterns observed in Fig 2: the 20% sticky cells correspond to isolated islands of sticky cells; the 30% sticky cells correspond to short filaments; and the 100% sticky cells correspond to clumps of sticky cells . The fraction of sticky cells in the liquid differs from that on the surface ( Fig 3B ) . For the pure strategy the difference is small . The fraction of sticky cells in the liquid is lower than that on the surface , due to the high cell division rate of non-sticky cells . Sticky cells that migrate to the surface remain attached until they die . Despite this permanent attachment , there is still a relatively large fraction of sticky cells in the liquid ( especially for high R values ) , because surface-attached sticky cells can dislodge their daughter cells to the liquid after cell division ( we assume that the colony is flat , so any cell division in the z-direction would result in the migration of a cell to the liquid; see Material and Methods ) . For the probabilistic strategy , the fractions of sticky cells in the liquid and on the surface are nearly the same . Even though there is a selective advantage for non-sticky cells in the liquid ( i . e . higher cell division rate ) , this does not affect the frequency of sticky cells too much , because all cells have a given probability to become sticky . For the decision-making strategy , we observed a surprisingly large difference between the fraction of sticky cells in the liquid and on the surface . At very low and very high costs of sticky cells ( R ≈ 1 or R ≈ 0 ) , the fraction of sticky cells in the decision-making strategy is more or less the same as that for the pure and probabilistic strategies , but at intermediate costs ( R = 0 . 2–0 . 8 ) almost 90% of the cells in the liquid are sticky . How can there be so many sticky cells in the liquid , while these cells have a lower cell division rate than the non-sticky cells ? In order to answer this question , we have to examine the population dynamics . There is a migratory asymmetry between the liquid and surface , while cells in the liquid can only migrate to the surface when finding a vacant position , cells from the surface can always migrate to the liquid and furthermore dislodge cells to the liquid during cell division . The migration rate of cells to the liquid is therefore much higher than that of cells to the surface . The migratory asymmetry is even bigger when the fraction of sticky cells on the surface is low , because this offers fewer possibilities for non-sticky cells to adhere to the surface . The surplus of migrants to the liquid results in a much higher competitive pressure in the liquid than on the surface . Cells therefore profit if they can increase the probability of surface attachment . Sticky cells are more effective migrants than non-sticky cells . As a result , cells in the decision-making strategy evolved such that they express the sticky phenotype in the liquid . Not all the replicate simulations evolved a high fraction of sticky cells in the liquid , because mutations that trigger cell differentiation in the liquid are often harmful on the surface ( S3 Fig ) . The pure and probabilistic strategies do not have a high fraction of sticky cells in the liquid , because cells cannot adjust their behaviour with respect to the environment in which they occur . Another surprising result in the decision-making strategy occurs at high costs of being sticky . At R = 0 , there are almost no sticky cells in the liquid , while approximately 20% of the cells on the surface are sticky ( see black arrows in Fig 3 ) . The lack of sticky cells in the liquid is surprising for two reasons . First , non-sticky cells cannot colonize the surface by themselves . If we would initiate our simulations with this evolved genotype , it would not be able to colonize the surface . Second , the surface contained around 2000 sticky cells . How can the number of sticky cells be so high , while sticky cells cannot divide and there are no sticky cells that migrate from the liquid to the surface ? In the next section , we address the above questions by examining the decision-making strategy in more detail . As shown by Fig 2 , in the decision-making strategy , at R = 0 , sticky cells are only surrounded by non-sticky neighbours . That means that a cell only differentiates when it has no sticky neighbours . If one of the neighbours is already sticky , a cell would remain non-sticky . Given this differentiation program , a colony would only be able to expand when the following sequence of fortunate events occurs ( see scheme in Fig 4 ) . First , the sticky cell should die . Second , in response to its death , some of the non-sticky neighbours should become sticky . At least two cells need to differentiate for the colony to split in two . Moreover , this should happen before the cells are dislodge from the surface , while in the absence of a sticky cell , non-sticky cells cannot stay on the surface . Finally , the remaining non-sticky neighbours have to divide in order to fill up the vacant positions next to the two novel sticky cells . Since colony fission starts with cell death , higher death rates of the sticky cells should increase the probability of colony fission . We examine this by competing three genotypes that have the same decision-making strategy , but differ with respect to the death rate of the sticky cells: the sticky cells have a 0% , 5% or 10% probability to die , respectively ( in the evolutionary simulations we assumed a 10% death probability ) . Hundred cells of each genotype were randomly placed on the surface . These genotypes were competed for 10 . 000 time steps under a reduced migration rate ( Pm = 0 . 01 ) , in order to focus on colony expansion . Fig 4 shows the frequency of each genotype over time . The genotype with the highest death rate indeed expanded faster ( green line in Fig 4 ) . This genotype had a competitive advantage over the other two genotypes at the onset of colony growth . However , when the population size went through its inflection point , the competitive advantage disappeared ( see black line in Fig 4 ) . At the inflection point , population growth is curtailed by the high cell density . There is less space to expand . As a consequence , colony longevity becomes more important for competition than colony expansion . Since lower death rates increase the longevity of a colony , the genotype with the lowest death rate slowly takes over the population of sticky cells ( see S2 Fig ) . Once all sticky cells belong to this genotype , there is no selective difference between the genotypes anymore , because the non-sticky cells of all genotypes are identical ( i . e . same fitness; see S2 Fig ) . In summary , Fig 4 illustrates that a single sticky cell , together with its non-sticky neighbours , can colonize the entire surface . The sticky cell and its neighbours often have the same genotype ( see time step 2000 in Fig 4 ) , because the non-sticky cells are produced by the sticky cell before cell differentiation or vice versa . Thus , cells inside the colony divide labour: sticky cells sacrifice their fitness , thereby increasing the fitness of their non-sticky clonal neighbours . Surface colonization in Fig 4 was characterized by two successional stages: the colonization stage and the climax stage . The colonizing genotype is favoured at low cell densities and the climax genotype at high cell densities . The successional stages in Fig 4 were based on differences in cell death ( note that in the evolutionary simulations the rate of cell death could not evolve and was kept constant ) . Similar successional stages might as well occur for different decision-making strategies . For example , in Fig 2 we observed two distinct decision-making strategies , each associated with a unique spatial pattern . One type consisted of isolated islands of sticky cells ( R = 0 and R = 0 . 4 ) and the other one of small filaments of sticky cells ( R = 0 . 8 ) . The isolated sticky cells were dominant at low R values and the filaments at high R values ( Figs 2 and 3 ) . Although the first colony type can expand over the surface , as shown in the previous section ( Fig 4 ) , there is still a substantial risk that all non-sticky cells are dislodged after the sticky cell dies . This risk is not present for the second colony type , because there are multiple concatenated sticky cells . Therefore we expected filamentous genotypes to be better colonizers . As in the previous section , we performed a competition experiment , placing hundred cells from each genotype–the isolated sticky cells and sticky filaments–on the surface . In order to focus on colony expansion , we reduced the migration rate from the liquid to the surface ( Pm = 0 . 01 ) . In addition , we assumed that sticky cells could not divide ( R = 0 ) . Fig 5 shows that filamentous sticky cells indeed function as a colonizing genotype , while the isolated sticky cells function as a climax genotype . The colonizing genotype shows a higher expansion rate than the climax genotype ( S4 Fig ) . However , when the population size goes through its inflection point , the fitness of the colonizing genotype drops and the climax genotype takes over . The climax genotype is less efficient in colonization , but–due to the lower fraction of sticky cells–has a higher cell division rate and therefore produces more propagules that can migrate to the liquid and initiate new colonies . In the evolutionary simulations , the climax genotype dominates the surface when the costs of being sticky are high ( i . e . low R values ) . Only when the costs of being sticky are fairly low ( R = 0 . 8 in Fig 2 and Fig 3 ) , the colonizing genotype ( i . e . filamentous genotype ) can outcompete the climax genotype , because the benefit of colony expansion outweigh the loss of propagule production . Since the colonizing genotype is more effective in colonizing the surface , the cell density at the surface is higher at low costs ( i . e . there is a sudden increase in the cell density around R = 0 . 7 in Fig 3A , which corresponds to the dominance of the colonizing genotype ) . Fig 5 illustrates that there is a trade-off between colony expansion and propagule production in our model: isolated sticky cells produce many propagules and expand slowly , while sticky filaments produce few propagules and expand rapidly . As a consequence , genotypes can specialize to grow at different stages of ecological succession . The colony’s expansion rate , longevity and propagule production depend on spatial pattern formation and , hence , the cell differentiation program that underlies phenotypic heterogeneity . In the previous sections , we investigated how the relative growth rate of sticky cells ( R ) affects the evolution of phenotypic heterogeneity . One would expect that ecological parameters play an important role as well . In this section , we vary both the relative cell division rate of sticky cells ( R ) and the migration rate towards the surface ( Pm ) . Note that migration towards the surface does not guarantee attachment , because cells first have to find an available spot , before they can actually attach . For each parameter combination we examine the evolved populations of all three differentiation strategies . We first examined the population size and fraction of sticky cells for each parameter combination of R and Pm ( R = 0–1 and Pm = 0–0 . 5 ) . This was done in both the liquid and on the surface . In the pure and probabilistic strategy , the population size and fraction of sticky cells were more or less independent of the migration rate in both the liquid and on the surface ( Fig 6 ) . As shown by Fig 3 , both the population size and the fraction of sticky cells decreased with the costs of being sticky ( Fig 6 ) . In the decision-making strategy , there was an effect of the migration rate , but only at high costs of being sticky . At R < 0 . 2 , higher migration rates towards the surface paradoxically lead to lower population densities on the surface ( Fig 6 ) . At the same time , there is also a change in the fraction of sticky cells in the liquid . At low migration rates the fraction of sticky cells in the liquid is nearly zero , while at high migration rates it is almost one . Sticky cells are more likely to colonize the surface than non-sticky cells , which makes it even more surprising that high migration rates result in a drop of the population density on the surface . How can we explain these paradoxical results ? One possible explanation for the low population density at high migration rates is the occurrence of exploitation . At high migration rates , non-sticky cells are more likely to migrate to the surface and exploit sticky cells . Such exploitation can lead to the collapse of colonies . To investigate whether or not such exploitation indeed occurs , we examined the evolutionary outcome of the decision-making strategy in more detail . For each parameter combination ( R = 0–1 and Pm = 0–0 . 5 ) , we examined the 25 most abundant genotypes that were present at the end of evolution . The decision-making strategy was determined for each genotype . That is , we determined for which conditions a cell would differentiate to a sticky cell on the surface and for which conditions it would differentiate in the liquid . On the surface , there are eight potential differentiation strategies ( S5 Fig ) : a cell could never or always differentiate or it could differentiate depending on the number of sticky neighbours ( with 6 potential thresholds ) . Each parameter combination was dominated by a particular decision-making strategy that was associated with a particular life cycle ( S5 Fig ) . Fig 7 shows the four dominant life cycles that evolved in the presence of phenotypic heterogeneity ( we ignored R = 1 and Pm = 0 , in which there was no heterogeneity ) . The life cycles consist of two stages: the colony stage at the surface and the propagule stage in the liquid . Spatial pattern formation influenced the colony properties and temporal pattern formation determined the colony’s life cycle . Colonies could reproduce in two ways: propagule production and colony fission . In life cycle 1 , cells formed filamentous colonies , like those shown above ( Figs 2 and 7 and S6 Fig ) . Filaments allow for efficient colony expansion . Colonies reproduce by filament breakage , mediated by cell death or cell differentiation ( i . e . colony fission ) , which results in two or more isolated filaments of sticky cells . At the same time , colonies reproduce by propagule production . Propagules get dislodged from the surface and migrate to establish new colonies in other regions of the surface . In life cycle 2 , at lower R values , colonies lose their filamentous property ( Figs 7 and S6 ) . The costs of being sticky outweigh the advantage of being filamentous . At this stage , the colonies consist of isolated sticky cells . Even though these colonies can still expand , via cell death ( Fig 4 ) , they expand slower than the filamentous colonies . The unoccupied parts of the surface leave more space for migrating propagules to establish new colonies ( Figs 6 and S6 ) . In life cycle 3 , at very low R values , there is a change in the propagule stage of the life cycle . Colonies still reproduce through fission , but propagules do not differentiate to sticky cells . Instead , they remain non-sticky and colonize the surface by exploiting the sticky cells . The non-sticky cells migrate to vacant positions that are available in existing colonies , such position are more common at low R values . Thus , the non-sticky migrants act as parasites that , in some cases , even take over established colonies ( see S3B Fig that shows how parasitizing propagules invade in the population ) . In life cycle 4 , at higher migration rates , colony formation becomes less common ( S6 Fig ) . At high migration rates , sticky cells on the surface are more likely to be exploited by non-sticky migrants from the liquid . This exploitation breaks down the benefits of colony formation . As a consequence , a new unicellular life cycle evolves . Cells differentiate to sticky cells in the liquid , which allows for surface colonization . Once at the surface , these sticky cells de-differentiate to non-sticky cells , which can divide before being dislodged to the liquid . In this life cycle , sticky-cells can still be exploited by non-sticky migrants , but it is less likely to occur , because cells are only sticky for a transient life stage . In the absence of multicellular colony formation , there is no colony expansion and , hence , a lower cell density on the surface . This explains why higher migration rates result in lower cell densities ( Fig 6 ) . In summary , at low costs of being sticky , colonies form filaments and reproduce by colony fission and propagule production ( Fig 7 ) . At intermediate costs , filamentous growth disappears and more space becomes available for surface colonization of propagules . At low costs , these propagules can only colonize the surface by parasitizing already existing colonies . At low costs and high migration rates , colony reproduction fails , because colonies succumb under the parasite pressure . In this case , a unicellular life cycle evolves in which surface attachment forms a transient life stage . Interestingly , the same life cycles also evolve for alternative surface geometries ( e . g . triangular and square grid , instead of hexagonal grid ) and for a three dimensional implementation of the surface ( see S2 Text and S7–S12 Figs ) . Only , when a cell’s neighbourhood is discontinuous–meaning that a cell’s neighbours do not neighbour each other–surface colonization becomes impossible at low R values ( see S2 Text ) .
Inside colonies , bacterial cells often express many different phenotypes . Using individual-based simulations , we studied the evolution of a sticky cell type in the context of surface colonization . We show that under the majority of parameter conditions surface colonization evolves . In many cases , colonization is associated with phenotypic heterogeneity , in which sticky and non-sticky cells co-occur on the surface . Phenotypic heterogeneity results from the trade-off between cell division and surface attachment ( see also [50–54] ) : sticky cells have a reduced cell division rate , but can colonize the surface , while non-sticky cells have a high cell division rate , but cannot colonize the surface by themselves . In our model , we compared three alternative differentiation strategies: pure strategy , probabilistic strategy and decision-making strategy . In the pure strategy , cells consistently express the same phenotype and can only switch via mutations . In the probabilistic strategy , cells differentiate with a certain probability . In the decision-making strategy , cells differentiate in response to the environment . Both the probabilistic and decision-making strategy evolve surface colonization for relatively high costs of being sticky , but only the latter can colonize the surface for extreme costs–i . e . when sticky cells hardly divide ( R << Pd ) . In the decision-making strategy , cells cooperate by dividing labour: the sticky cell sacrifices its fitness for the benefit of the colony ( i . e . non-sticky cells that surround the sticky cell ) . Cells in the probabilistic strategy can reciprocate benefits , but are not capable of coordinating their behaviour . One striking outcome of the model is that–under the decision-making strategy–different life cycles evolved [55] . The evolution of life cycles follows from the dynamical environment in which bacterial cells live , in which they can alternate between growing on the surface and growing in the liquid . In most life cycles , the surface-attached colony forms the dominant life stage . The surface provides a scaffold on which cells can organize themselves [56 , 57] . The propagules in the liquid have only a marginal chance to colonize the surface . Yet , once a propagule attaches to the surface , it can form a colony that is relatively long lived and thereby produces many new propagules . Cells can affect the colony properties by coordinating their behaviour . The fitness of a colony is determined by three key properties [55]: ( 1 ) colony expansion , ( 2 ) colony longevity and ( 3 ) propagule production . Cells increase the longevity of colonies by regulating cell differentiation: if the sticky cell dies one or more neighbouring cells will differentiate , thereby guaranteeing the survival of the colony . This simple type of coordination allows for surface colonization in the toughest conditions ( R = 0 ) . Life cycles can only evolve in the decision-making strategy , because spatial and temporal organization can only come about when cells can respond to the environment [58–61] . For simplicity , we assumed that cells could only sense two environmental cues , yet , in reality cells can sense many more cues [62] , which presumably allow for many alternative forms of coordination . It would be interesting to explore how environmental information facilitates or constrains spatial and temporal organization . The surface-attached colony is vulnerable to exploitation . Cells inside the colony cooperate . As explained above , the sticky cell sacrifices its fitness for the benefit of the colony . As long as the sticky cell is surrounded by its non-sticky siblings , it cannot be exploited . However , when one of the neighbours dies , there is a risk that the cell from the liquid migrates next to the sticky cells and reaps the benefits of surface attachment without paying the costs . In our model , exploitation is more likely to occur when migration rate is high , when a cell’s neighbourhood is large and when a cell’s neighbourhood is discontinuous ( see S2 Text ) . More generally , one could say that adhesive cells are less likely to be exploited when the diffusion of adhesive molecules is limited , such that only sibling cells profit from adhesion and invasion from outside is minimized [34 , 35 , 38 , 39 , 63–67] . A recent study of Nadell and colleagues [68] also illustrates that cells can actively prevent exploitation . They show that in Vibrio cholerae colonies , cells secrete a protein that facilitates a closer association between cells and the extracellular matrix . This prevents open spaces in the colony , which subsequently prevents cells from invading the interior of the colony and hence exploitation . Alternatively , non-adhesive cells might simply be less likely to join a colony than adhesive cells [69 , 70] . This results in the segregation of adhesive and non-adhesive cells , which in turn prevents exploitation . Bacterial life cycles are often hard to study empirically . First , it is nearly impossible to trace bacterial individuals in nature . Second , in many cases , the ecological relevant unit of a bacterial life cycle is not the individual cell , but the colony [58 , 71] . Cells are relatively short-lived an only survive a part of colony formation , yet colonies often go through coordinated life stages [1 , 2 , 72] . Thus , instead of tracing the fate of a single cell , one should trace the fate of all cells in the colony . Despite these difficulties , life cycles are well characterized for a number of bacterial species [73] . For many of them , the life cycle consists of a surface-attached life stage and a unicellular dispersal stage [61 , 74] . In the surface-attached life stage , cells often organize into colony structures that facilitate colony expansion or dispersal . For example , many bacteria develop filamentous structures to facilitate colony expansion [75–78] , while other bacteria develop fruiting bodies to facilitate dispersal [79–84] . Our study indicates that the ecological significance of the observed colony structures–and the associated adhesive cell types–can only be fully appreciated when considering the entire life cycle of a bacterium , including the dynamical environment in which bacteria make their living .
Cells were grown in 2 . 5mL static liquid MSgg [85] using twelve-well plates . Plates were incubated for 50 hours at 30°C . The inoculum was prepared by growing strains overnight on 1 . 5% agar LB plates at 37°C . Overnight colonies were scraped from the plates and diluted in phosphate buffered saline ( PBS ) to an optical density of 0 . 2 ( OD600 = 0 . 2 ) . The wells were inoculated with 2μL of this sample . All strains were derived from a non-domesticated wild type B . subtilis strain called NCIB 3610 [86] . The regulatory mutant strain could not express two operons , eps and tapA , which are essential for matrix production ( strain DS91 , see [85 , 87] ) . The fluorescent strain expresses cyan fluorescent protein ( CFP , artificially coloured red in S1 Fig ) in all matrix-producing cells ( strain DL823 , see [16] ) . For microscopy , cells were isolated from top of the pellicle ( i . e . colony at air-liquid interface ) , placed on an object glass with solidified 200μL of 1 . 5% agarose PBS and examined using an inverted microscope . The inverted microscope was a Nikon Eclipse TE2000-U microscope equipped with a 20× Plan Apo objective and a 60× Plan Apo oil objective . Images were taken using CFP filter . Image analysis was performed with ImageJ ( 1 . 48v ) . In the model we assume there are two niches: the surface and the liquid . At the onset of evolution cells only occur in the liquid , where cells are assumed to freely float around and there is no spatial structure . The surface consists of a hexagonal grid ( Fig 1 ) . In order to colonize the surface , cells have to become ‘sticky’ . When cells are sticky they can attach to the surface . Sticky cells not only facilitate their own attachment to the surface , they can also mediate non-sticky cells to adhere to the surface , but only if these non-sticky cells are located immediately adjacent to the sticky cells . Since the surface consists of a hexagonal grid , one-sticky cell can be surrounded by maximally six non-sticky cells . Stickiness typically results from the production of costly substances , such as extracellular polysaccharides , we therefore assume that being sticky reduces the rate of cell division ( R ) . Yet , despite these costs , becoming sticky can be beneficial , while cells can avoid competition in the liquid by adhering to the surface . Once adhered to the surface , a cell can also migrate back to the liquid , by de-differentiating to a non-sticky cell . At each time step , one out of five events can occur: ( 1 ) cell migration from the liquid to the surface , ( 2 ) cell migration from the surface to the liquid , ( 3 ) cell differentiation , ( 4 ) cell death and ( 5 ) cell division . For each cell , the event that occurs is selected randomly , to randomize the order in which cell events occur . We tested three different version of the model in which the strategy underlying cell differentiation is different ( Fig 1 ) . These strategies will be discussed in detail below . | In nature , most bacteria occur in surface-attached colonies . Inside these colonies , cells often express many different phenotypes . The significance of these phenotypes often remains unknown . We study the evolution of cell differentiation in the context of surface colonization . We particularly focus on the evolution of a ‘sticky’ cell type that is needed for surface attachment . We show that the sticky cell type readily evolves and escapes from competition in the liquid by attaching to the surface . In most cases , surface colonization is accompanied by phenotypic heterogeneity , in which sticky and non-sticky cell co-occupy the surface . The non-sticky cells hitchhike with the sticky cells , thereby profiting from surface attachment without paying the cost of being sticky . In the presence of regulation , cell differentiation leads to the evolution of intricate bacterial life cycles in which cells alternate between living in surface-attached colonies and living in the liquid . The bacterial life cycles are orchestrated by temporal and spatial pattern formation of cell types . Our model illustrates how cell differentiation can be of key importance for the evolution of bacterial life cycles . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"cell",
"death",
"cell",
"motility",
"organismal",
"evolution",
"decision",
"making",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"microbiology",
"neuroscience",
"cell",
"differentiation",
"developmental",
"biology",
"cognition",
"mi... | 2016 | Phenotypic Heterogeneity and the Evolution of Bacterial Life Cycles |
Regulation of synaptic AMPA receptor levels is a major mechanism underlying homeostatic synaptic scaling . While in vitro studies have implicated several molecules in synaptic scaling , the in vivo mechanisms linking chronic changes in synaptic activity to alterations in AMPA receptor expression are not well understood . Here we use a genetic approach in C . elegans to dissect a negative feedback pathway coupling levels of the AMPA receptor GLR-1 with its own transcription . GLR-1 trafficking mutants with decreased synaptic receptors in the ventral nerve cord ( VNC ) exhibit compensatory increases in glr-1 mRNA , which can be attributed to increased glr-1 transcription . Glutamatergic transmission mutants lacking presynaptic eat-4/VGLUT or postsynaptic glr-1 , exhibit compensatory increases in glr-1 transcription , suggesting that loss of GLR-1 activity is sufficient to trigger the feedback pathway . Direct and specific inhibition of GLR-1-expressing neurons using a chemical genetic silencing approach also results in increased glr-1 transcription . Conversely , expression of a constitutively active version of GLR-1 results in decreased glr-1 transcription , suggesting that bidirectional changes in GLR-1 signaling results in reciprocal alterations in glr-1 transcription . We identify the CMK-1/CaMK signaling axis as a mediator of the glr-1 transcriptional feedback mechanism . Loss-of-function mutations in the upstream kinase ckk-1/CaMKK , the CaM kinase cmk-1/CaMK , or a downstream transcription factor crh-1/CREB , result in increased glr-1 transcription , suggesting that the CMK-1 signaling pathway functions to repress glr-1 transcription . Genetic double mutant analyses suggest that CMK-1 signaling is required for the glr-1 transcriptional feedback pathway . Furthermore , alterations in GLR-1 signaling that trigger the feedback mechanism also regulate the nucleocytoplasmic distribution of CMK-1 , and activated , nuclear-localized CMK-1 blocks the feedback pathway . We propose a model in which synaptic activity regulates the nuclear localization of CMK-1 to mediate a negative feedback mechanism coupling GLR-1 activity with its own transcription .
Homeostatic synaptic plasticity alters synaptic strengths in order to compensate for perturbations in neuronal activity . Homeostasis is thought to stabilize neuronal firing rates to remain within a physiological range in response to developmental changes in connectivity or alterations in synaptic strength during experience-dependent plasticity [1 , 2] . Synaptic scaling is a form of homeostatic synaptic plasticity that has been widely studied in vitro [2–6] and in vivo after sensory deprivation in the rodent visual cortex [7–9] . One major mechanism underlying changes in synaptic strength during synaptic scaling is the regulation of AMPA receptor ( AMPAR ) levels at synapses . During homeostatic scaling , chronic activity-blockade or enhancement of activity results in compensatory increases or decreases , respectively , in AMPAR abundance at synapses . These changes in synaptic AMPARs are achieved , in part , by altering the rates of receptor exo- or endocytosis [3–6 , 10–14] . Many molecules have been implicated in regulating synaptic AMPAR levels during homeostasis [11–13 , 15–17] . In particular , homeostatic synaptic plasticity requires calcium signaling and the CaM kinases CaMKK and CaMKIV [3 , 18–20] . Inhibition of calcium transients or CaMK signaling phenocopies activity-blockade and leads to increases in synaptic AMPARs [19] . Similarly , inhibition of voltage-gated calcium channels or CaMK signaling prevents scaling down of synaptic AMPARs [18] . Homeostatic synaptic plasticity is dependent on transcription , as pharmacological inhibition of transcription prevents bidirectional synaptic scaling [18 , 19 , 21 , 22] . Interestingly , activity-blockade results in decreased levels of activated CaMKIV in the nucleus in a transcription-independent manner [19] , suggesting that CaMKIV may translocate between the cytoplasm and nucleus during synaptic scaling to regulate transcription . These studies suggest that nuclear CaMKIV represses synaptic scaling and the associated increase in synaptic AMPARs in response to activity-blockade , but the transcriptional targets of CaMKIV responsible for the increase in synaptic AMPARs have not been defined . Here we investigate a compensatory feedback pathway in C . elegans where synaptic levels of the AMPAR GLR-1 are negatively coupled to glr-1 transcription via the CMK-1/CaMK signaling pathway . In C . elegans , CMK-1 is the sole ortholog of mammalian CaMKI and CaMKIV . As in mammals , CMK-1 is phosphorylated by CKK-1/CaMKK and can regulate CRH-1 , the C . elegans homolog of CREB [23–25] . Recent studies in C . elegans show that CMK-1 can shuttle between the nucleus and cytoplasm to regulate temperature thresholds and experience-dependent thermotaxis under physiologic temperature and in response to noxious heat [26–28] . While much progress has been made identifying molecules involved in homeostatic synaptic scaling in neuronal and slice cultures [13] , in vivo studies of mechanisms directly linking chronic changes in activity to regulation of AMPAR expression are lacking . Here we use a genetic approach to identify in vivo mechanisms involved in a negative feedback pathway in C . elegans that is reminiscent of synaptic homeostasis . We show that chronic activity-blockade or enhancement of GLR-1 function results in bidirectional changes in glr-1 transcription in vivo . We find that regulation of glr-1 transcription in response to chronic changes in synaptic activity requires the CMK-1 signaling pathway and redistribution of CMK-1 between the nucleus and cytoplasm . This study identifies the signaling mechanism underlying a compensatory feedback pathway that couples GLR-1 with its own transcription .
We previously found that trafficking mutants with reduced GLR-1 abundance at synapses in the ventral nerve cord ( VNC ) exhibit reciprocal increases in glr-1 mRNA levels . Specifically , animals with mutations in the deubiquitinating enzyme USP-46 , which removes ubiquitin from GLR-1 and protects it from degradation , exhibit decreased levels of GLR-1 in the VNC and a compensatory 3 fold increase in glr-1 transcript levels as measured by real-time quantitative PCR ( RT-qPCR ) [29] . Similarly , mutations in the kinesin motor KLP-4/KIF13 , which positively regulates GLR-1 trafficking to the VNC , result in decreased levels of GLR-1 in the VNC and a compensatory 2–3 fold increase in glr-1 transcript levels [30] . We hypothesized that GLR-1 levels or function at synapses in the VNC are monitored and coupled via a negative feedback mechanism to glr-1 transcript levels . To investigate the molecular mechanisms involved in this feedback pathway , we created a series of transgenic animals expressing different combinations of a nuclear-localized GFP reporter ( NLS-tagged GFP fused to LacZ ) under control of the glr-1 promoter ( Pglr-1 ) and/or the glr-1 3’ untranslated region ( UTR ) . Pglr-1 includes 5 . 3 kilobases of sequence upstream of the transcription start site [31] and allows monitoring of transcriptional activity of the promoter . The glr-1 3’UTR includes 100 base pairs downstream of the ORF , as predicted by modENCODE [32] , and allows us to monitor the contribution of the 3’UTR to transcript levels . We first validated this glr-1 reporter under control of both Pglr-1 and the glr-1 3’UTR by testing if GFP fluorescence was altered in klp-4/KIF13 trafficking mutants . Briefly , we measured the maximum fluorescence intensity of GFP in the nucleus of the GLR-1-expressing interneuron PVC in wild type and klp-4 ( tm2114 ) loss-of-function mutants ( see Materials and Methods ) . We found that GFP fluorescence increased in klp-4 ( tm2114 ) mutants ( Fig 1A ) , consistent with our previous RT-qPCR results [30] . Because klp-4 mutants have reduced GLR-1 at synapses in the VNC , this data implies that decreased synaptic GLR-1 may trigger a compensatory feedback pathway resulting in increased glr-1 transcript . To directly test if loss of GLR-1 itself could trigger the feedback pathway , we measured the GFP reporter under control of Pglr-1 and the glr-1 3’UTR in glr-1 ( n2461 ) null mutants . We found that GFP fluorescence increased in glr-1 mutants to a similar extent as in klp-4 mutants ( Fig 1A and 1B ) . These data suggest that decreased GLR-1 protein or function is sufficient to trigger a compensatory feedback mechanism negatively coupling GLR-1 to its own transcript levels . These data also indicate that the glr-1 promoter together with the glr-1 3’UTR are sufficient to mediate the feedback mechanism . To determine the respective contributions of Pglr-1 and the glr-1 3’UTR to the feedback mechanism , we generated additional GFP reporter transgenes consisting of NLS-GFP-LacZ under the control of either the glr-1 or nmr-1 promoters combined with either the glr-1 or unc-54 3’UTRs . The nmr-1 promoter provides an alternative promoter that is expressed in an overlapping set of neurons with GLR-1 , including the interneuron PVC [33] . The unc-54 3’UTR is widely used for permissive gene expression in C . elegans [34] . We crossed these GFP reporter transgenes into several genetic backgrounds and measured GFP fluorescence in the nucleus of PVC interneurons as described above . When fluorescence was measured from a GFP reporter under control of the nmr-1 promoter ( Pnmr-1 ) and the glr-1 3’UTR , we observed no significant change in fluorescence in either klp-4 ( tm2114 ) or glr-1 ( n2461 ) loss-of-function mutants ( Fig 1C and 1D ) . This result suggests that the glr-1 3’UTR is not sufficient to mediate the feedback mechanism . On the other hand , when GFP fluorescence was measured from the reporter transgene containing Pglr-1 and the unc-54 3’UTR ( hereafter referred to as the glr-1 transcriptional reporter ) , we observed a small but significant increase in fluorescence in both klp-4 and glr-1 mutants ( Fig 1E and 1F ) . This glr-1 transcriptional reporter was also increased in usp-46 ( ok2232 ) loss-of-function mutants ( Fig 1G ) , consistent with our previous RT-qPCR results [29] . Importantly , the nmr-1 promoter and the unc-54 3’UTR are not regulated by the feedback pathway because a GFP reporter containing these elements was unaltered in klp-4 and glr-1 mutants ( S1 Fig ) . Together , these data indicate that Pglr-1 is sufficient to mediate the feedback mechanism , suggesting that neurons respond to decreased GLR-1 levels or function in the VNC by increasing glr-1 transcription . We next investigated whether the feedback mechanism was bidirectional by testing if increased GLR-1 in the VNC triggers a decrease in glr-1 transcription . UNC-11/AP180 is a clathrin adaptin involved in endocytosis of GLR-1 , and the receptor accumulates at the plasma membrane in the VNC of unc-11 mutants [35] . We found that fluorescence of the GFP reporter under control of Pglr-1 and the glr-1 3’UTR decreased in unc-11 ( e47 ) null mutants ( Fig 1H ) . We observed a similar reduction of the glr-1 transcriptional reporter in unc-11 mutants ( Fig 1I ) , suggesting that Pglr-1 is sufficient to mediate decreased glr-1 transcription . Interestingly , genetic double mutant analyses indicate that the effects of unc-11 on glr-1 transcription are not dependent on glr-1 ( S2 Fig ) . Together , these data suggest that mutation of the clathrin adaptin unc-11/AP180 likely blocks the endocytosis of another membrane protein or ion channel in addition to GLR-1 , whose accumulation results in reduced glr-1 transcription . We performed several experiments to test if changes in glutamate signaling , rather than levels of synaptic GLR-1 , were sufficient to trigger the transcriptional feedback mechanism . First , we tested whether reductions in glutamatergic transmission could trigger the feedback mechanism by analyzing glr-1 expression in eat-4 synaptic transmission mutants . EAT-4 is a vesicular glutamate transporter ( VGLUT ) responsible for loading glutamate into synaptic vesicles [36 , 37] . Loss of eat-4 results in defects in glutamatergic transmission [37 , 38] and a compensatory increase in synaptic GLR-1 in the VNC [10] . We found that eat-4 ( n2474 ) loss-of-function mutants exhibit increased endogenous glr-1 mRNA levels compared to wild type controls using RT-qPCR ( Fig 2A ) . In support of this data , we found that eat-4 ( n2474 ) mutants also exhibit increased GFP fluorescence from the reporter under control of Pglr-1 and the glr-1 3’UTR ( Fig 2B ) . Furthermore , Pglr-1 was sufficient to mediate this effect because GFP fluorescence still increased in eat-4 ( n2474 ) mutants expressing the glr-1 transcriptional reporter ( Fig 2C ) . Together , these data suggest that chronic decreases in glutamate signaling ( Fig 2 ) or postsynaptic glutamate receptors ( Fig 1 ) are sufficient to trigger the glr-1 transcriptional feedback pathway . We next investigated whether direct and more acute suppression of neuronal activity specifically in GLR-1-expressing neurons could trigger the feedback mechanism using a recently developed chemical genetic silencing strategy . Ectopic expression of a Drosophila histamine-gated chloride channel ( HisCl1 ) in C . elegans neurons enables relatively acute repression of neuronal activity by exogenous histamine [39] . We generated transgenic animals expressing HisCl1 in GLR-1-expressing neurons ( Pglr-1::HisCl1 ) and verified the efficacy of this silencing approach by measuring GLR-1-dependent locomotion reversal behavior . The frequency of spontaneous reversals is regulated by glutamatergic signaling , and mutants with reduced glutamatergic signaling ( i . e . , glr-1 or eat-4 mutants ) exhibit decreased reversal frequencies [33 , 35 , 40] . We found that exposure of animals expressing HisCl1 to exogenous histamine for 10 minutes led to a dramatic decrease in spontaneous reversal frequency compared to wild type controls ( Fig 2D ) . This data suggests that activation of HisCl1 channels specifically in GLR-1-expressing neurons suppresses their activity and impacts GLR-1-dependent locomotion behavior . In order to test whether direct inhibition of GLR-1-expressing neurons could increase glr-1 transcription , we exposed HisCl1-expressing animals to histamine for one and four hours and then measured Pglr-1 activity using the glr-1 transcriptional reporter . Fluorescence at each time point was normalized to HisCl1-expressing animals in the absence of histamine ( see Materials and Methods ) . We found a small increase in GFP reporter fluorescence after both one and four hours of histamine treatment ( Fig 2E ) . Although the histamine-induced effect on the glr-1 transcriptional reporter was modest , it was significant ( p<0 . 05 ) and suggests that direct inhibition of GLR-1-expressing neurons can trigger an increase in glr-1 transcription . In contrast , wild type animals not expressing HisCl1 showed no significant increase in Pglr-1 activity when exposed to histamine ( S3 Fig ) . We did , however , observe a reduction in Pglr-1 activity in wild type animals after four hours of histamine exposure ( S3 Fig ) . Unfortunately , this decrease in Pglr-1 activity precluded our ability to test whether long term inhibition by histamine could also induce a late glr-1 transcriptional response . Nevertheless , these results suggest that decreasing neuronal activity specifically in GLR-1-expressing neurons can trigger the feedback mechanism to increase glr-1 transcription in the mature nervous system . Finally , we investigated whether directly increasing GLR-1 function could regulate the transcriptional feedback pathway . We increased GLR-1 activity in a subset of interneurons by expressing a mutant version of GLR-1 ( under control of the nmr-1 promoter ) , that contains an alanine to threonine substitution ( A/T ) in the pore domain resulting in a constitutively active channel with increased conductance [40] . Animals expressing GLR-1 ( A/T ) exhibit a dramatic increase in spontaneous locomotion reversals consistent with increased glutamatergic signaling [29 , 40 , 41] . We found that Pglr-1 activity decreased in GLR-1 ( A/T ) -expressing animals compared to wild type controls ( Fig 2F ) . These data are consistent with the hypothesis that increased GLR-1 function triggers the feedback pathway to reduce glr-1 transcription . Together , our data show that increasing or decreasing glutamatergic signaling results in compensatory and reciprocal changes in glr-1 transcription . CaM kinases ( CaMKs ) I and IV are important mediators of calcium-dependent signaling mechanisms involved in neuronal development and function . In particular , CaMKIV can mediate activity-dependent regulation of gene transcription [42] , and has been shown to mediate AMPAR-dependent homeostatic synaptic scaling in a transcription-dependent manner [18 , 19] . In C . elegans , CMK-1 , the homolog of CaMKI and CaMKIV [24] , is widely expressed in the nervous system [26] , and has been shown to function in specific sensory neurons to mediate experience-dependent thermotaxis at physiological temperatures and avoidance of noxious heat [26–28] . However , the downstream transcriptional targets of CMK-1 and CaMKIV that mediate their physiological effects are not clear . To test whether CMK-1 was involved in regulating glr-1 transcription , we first measured endogenous glr-1 mRNA levels in cmk-1 ( oy21 ) loss-of-function mutants using RT-qPCR . Intriguingly , we found increased glr-1 mRNA levels relative to two reference genes ( act-1 and ama-1 ) in cmk-1 ( oy21 ) mutants ( Fig 3A ) , suggesting that CMK-1 negatively regulates glr-1 transcript levels . Consistent with this result , loss-of-function mutations in ckk-1/CaMKK , the upstream activator of CMK-1 , resulted in increased GFP fluorescence from a reporter under control of Pglr-1 and the glr-1 3’UTR ( Fig 3B ) . We next tested whether Pglr-1 was sufficient to mediate the effects of the CMK-1 pathway using the glr-1 transcriptional reporter . We found that Pglr-1 activity increased in ckk-1 ( ok1033 ) loss-of-function mutants ( Fig 3C ) and two independent loss-of-function alleles of cmk-1 ( oy21 and ok287 ) ( Fig 3D and 3E ) . These results indicate that the CMK-1 signaling pathway acts basally to repress glr-1 transcription . Expression of cmk-1 cDNA specifically in GLR-1-expressing neurons rescues the increase in Pglr-1 activity observed in cmk-1 ( oy21 ) loss-of-function mutants ( Fig 3H ) , whereas expression of a kinase-dead version of CMK-1 ( K52A ) [25] does not rescue ( Fig 3I ) . The difference in rescue between the wild-type and kinase-dead versions of CMK-1 cannot be explained by different levels of transgene expression , as both Pglr-1::CMK-1 and Pglr-1::CMK-1 ( K52A ) transgenes were expressed at comparable levels as determined by RT-qPCR ( S4 Fig ) . These results indicate that CMK-1 functions in a kinase-dependent manner specifically in GLR-1-expressing neurons to repress glr-1 transcription . CaMKI and CaMKIV in mammals , and CMK-1 in C . elegans , have been shown to phosphorylate the transcription factor cyclic AMP response element binding protein ( CREB ) to regulate gene expression [24 , 25 , 43–46] . Thus , we tested whether mutations in crh-1 , the C . elegans homolog of CREB , affected glr-1 transcription . We found that fluorescence of the glr-1 transcriptional reporter was increased in crh-1 ( tz2 ) loss-of-function mutants ( Fig 3F ) , consistent with a role for CREB as a downstream target of CMK-1 in regulating glr-1 transcription . Additionally , since CREB is known to function together with the transcriptional co-activator CREB binding protein ( CBP-1 ) /p300 which can also be phosphorylated by CaMKIV [42 , 47] , we took advantage of a gain-of-function allele in cbp-1 ( ku258 gf ) [48] to test if cbp-1 was involved in regulating glr-1 transcription . We found that cbp-1 ( ku258 gf ) mutants exhibited decreased fluorescence of the glr-1 transcriptional reporter ( Fig 3G ) . Together , these results indicate that the CaMK signaling axis , including CKK-1/CaMKK , CMK-1/CaMK , CRH-1/CREB and CBP-1/CBP act to repress glr-1 transcription ( Fig 3J ) . To test whether the negative feedback pathway triggered by loss of glr-1 was mediated by CMK-1 signaling , we generated a series of genetic double mutants between glr-1 and various CMK-1 pathway mutants . We hypothesized that if decreased GLR-1 signaling triggers an increase in glr-1 transcription by deactivating the CMK-1 pathway , we would expect glr-1; cmk-1 double mutants to have non-additive effects on the glr-1 transcriptional reporter . Alternatively , if cmk-1 functions in an independent pathway to regulate glr-1 transcription , we would expect glr-1; cmk-1 double mutants to have an additive effect on the glr-1 transcriptional reporter . We found that glr-1 ( n2461 ) ; cmk-1 ( oy21 ) double mutants exhibited an increase in the glr-1 transcriptional reporter that is indistinguishable from either single mutant ( Fig 4A ) . This non-additive effect is consistent with the idea that the glr-1-triggered feedback mechanism and cmk-1 function in the same pathway to increase glr-1 transcription . In support of this finding , we found that glr-1 ( n2461 ) ; crh-1 ( tz2 ) double mutants also exhibit an increase in the glr-1 transcriptional reporter that was identical to either single mutant ( Fig 4B ) , suggesting that CRH-1/CREB also likely functions in the same pathway to negatively regulate glr-1 transcription . Although these non-additive effects support the idea that the CMK-1 pathway may mediate the glr-1 transcriptional feedback mechanism , we cannot formally rule out a potential ceiling effect of the reporter . To provide further genetic evidence for a role for CMK-1 in the glr-1 transcriptional feedback mechanism , we tested whether a recently isolated gain-of-function ( gf ) allele of cmk-1 , pg58 , could suppress the increase in glr-1 transcription observed in glr-1 mutants . cmk-1 ( pg58 gf ) contains a premature stop codon at W305 resulting in a truncated version of CMK-1 ( 1–304 ) . CMK-1 ( 1–304 ) is missing most of its regulatory domain and a putative nuclear export sequence ( NES ) , and the altered protein has been shown to accumulate in the nucleus [27] . Interestingly , we found that although cmk-1 ( pg58 gf ) did not affect basal glr-1 transcription , this gain-of-function allele completely blocked the increase in the glr-1 transcriptional reporter triggered by loss of glr-1 ( Fig 4C ) . Together , these data are consistent with the model that CMK-1 signaling mediates the glr-1 transcriptional feedback mechanism . CaM kinases are well-known mediators of activity-dependent gene expression , and specific isoforms have been shown to translocate between the cytoplasm and nucleus [42 , 49] . For example , in mammalian neuronal cultures , homeostatic increases in synaptic GluRs are correlated with a reduction in activated CaMKIV in the nucleus [19] . In C . elegans , CMK-1 can shuttle between the cytoplasm and nucleus to regulate thermosensory behaviors [27 , 28] . Thus , we tested whether alterations in glr-1 transcription were accompanied by changes in the subcellular localization of CMK-1 . We expressed GFP-tagged CMK-1 ( CMK-1::GFP ) [26] in GLR-1-expressing interneurons and used confocal microscopy to determine the relative subcellular localization of CMK-1::GFP in the cytoplasm versus nucleus of PVC neurons ( see Materials and Methods ) . The subcellular localization of CMK-1::GFP is regulated by changes in physiological temperature and noxious heat [27 , 28] , and CMK-1::GFP can rescue heat avoidance behavioral defects in cmk-1 mutants , suggesting that the tagged protein is functional [27] . Since CKK-1 phosphorylation of CMK-1 has been shown to promote the nuclear accumulation of CMK-1::GFP in sensory neurons [27 , 28] , we first analyzed the subcellular localization of CMK-1::GFP ( Fig 5A ) in GLR-1-expressing neurons in ckk-1 ( ok1033 ) loss-of-function mutants . We found that CMK-1::GFP decreases in the nucleus and increases in the cytoplasm in ckk-1 ( ok1033 ) mutants ( Fig 5B ) . In other words , the subcellular localization of CMK-1::GFP shifts from the nucleus towards the cytoplasm in ckk-1 mutants , consistent with previous studies [27 , 28] . To test whether the subcellular localization of CMK-1 is regulated by GLR-1 signaling , we analyzed the distribution of CMK-1::GFP in glr-1 mutants . Similar to ckk-1 mutants , we found that CMK-1::GFP decreases in the nucleus and increases in the cytoplasm in glr-1 ( n2461 ) mutants ( Fig 5B ) . These results are consistent with the idea that decreased synaptic GLR-1 results in increased retention of CMK-1 in the cytoplasm and relief of repression of glr-1 transcription . In contrast , we found that increasing GLR-1 signaling by expression of constitutively active GLR-1 ( A/T ) in interneurons results in increased localization of CMK-1::GFP to the nucleus ( Fig 5C ) . Together , these data suggest that increased or decreased GLR-1 signaling in interneurons results in increased or decreased accumulation , respectively , of CMK-1 in the nucleus . To specifically test whether nuclear localization of CMK-1 is sufficient to repress the increase in glr-1 transcription triggered by loss of glutamatergic signaling , we expressed a constitutively nuclear-localized version of CMK-1 containing an exogenous NLS ( Pglr-1::CMK-1::EGL-13-NLS ) in GLR-1-expressing neurons . CMK-1::EGL-13-NLS was shown to be five-fold enriched in the nucleus where it can rescue cmk-1 null mutants for several thermosensory defects [28] . We found that expression of constitutively nuclear CMK-1 was sufficient to block the increase in the glr-1 transcriptional reporter observed in glr-1 ( n2461 ) mutants ( Fig 5D ) . These data suggest that nuclear localization of CMK-1 represses glr-1 transcription and provides further evidence that the CMK-1 signaling pathway mediates the glr-1 transcriptional feedback mechanism ( Fig 5E ) .
We found that GLR-1 trafficking mutants ( i . e . , klp-4/KIF13 or usp-46 mutants ) with decreased GLR-1 in the VNC exhibit compensatory increases in glr-1 expression ( Fig 1 ) . Analysis of fluorescent reporters containing either Pglr-1 or the glr-1 3’UTR revealed that the glr-1 promoter was sufficient to mediate the feedback mechanism ( Fig 1 ) . Interestingly , although the glr-1 3’UTR alone did not appear to be sufficient to mediate the feedback pathway ( Fig 1C and 1D ) , we noticed that reporter constructs containing the glr-1 3’UTR together with Pglr-1 ( Figs 1A , 1B , 1H and 3B ) appear to have larger magnitude effects versus the unc-54 3’UTR ( Figs 1E , 1F , 1I and 3C ) hinting at a potential contribution of the glr-1 3’UTR . Statistical comparison of the relevant data sets revealed significant contributions ( p<0 . 05 , Two-way ANOVA ) of the glr-1 3’UTR ( together with Pglr-1 ) in klp-4 ( p = 0 . 03 ) and ckk-1 ( p = 0 . 03 ) mutant backgrounds . The contribution of the glr-1 3’UTR versus the unc-54 3’UTR in glr-1 ( p = 0 . 1 ) and unc-11 ( p = 0 . 2 ) mutant backgrounds did not reach statistical significance . Thus , the glr-1 3’UTR appears to contribute to the regulation of glr-1 expression in the feedback pathway in some genetic backgrounds . A more detailed analysis of the glr-1 3’UTR together with other endogenous regulatory elements is warranted to fully understand the role of the glr-1 3’UTR in the feedback pathway . Interestingly , a recent study in rodent hippocampal neurons showed that microRNA miR-92A inhibits translation of GluA1 by binding to its 3’UTR , and that this miRNA-mediated mechanism regulates homeostatic scaling in response to chronic activity-blockade [50] . However , we did not find any conserved miRNA binding sites in the glr-1 3’UTR using several target site prediction algorithms . Furthermore , we found that the glr-1 3’UTR alone was not sufficient to mediate the feedback mechanism in C . elegans ( Fig 1C and 1D ) . Thus , while non-conserved miRNAs may still contribute to the regulation of the glr-1 3’UTR , this regulation does not appear to be sufficient to mediate the feedback pathway . We also investigated whether changes in glutamate signaling could trigger the feedback mechanism . We found that glutamatergic transmission mutants lacking glr-1 itself ( Fig 1 ) or the presynaptic eat-4/VGLUT ( Fig 2 ) were sufficient to trigger the glr-1 transcriptional feedback mechanism . Furthermore , expression of a constitutively active GLR-1 , GLR-1 ( A/T ) , resulted in decreased glr-1 transcription ( Fig 2F ) . These data indicate that bidirectional changes in GLR-1 signaling are negatively coupled to glr-1 transcription . A previous study showed that chronic activity-blockade in eat-4/VGLUT mutants results in a homeostatic compensatory increase in synaptic GLR-1 levels that is mediated by changes in clathrin-mediated endocytosis [10] . We found that eat-4 mutants also exhibit increased endogenous glr-1 transcript based on RT-qPCR and increased Pglr-1 activity based on a glr-1 transcriptional reporter expressing nuclear-localized NLS-GFP-LacZ ( Fig 2 ) . Given the multiple mechanisms that contribute to synaptic scaling in mammalian neurons , we suspect that the homeostatic compensatory increase in GLR-1 observed in eat-4 mutants is likely mediated by several mechanisms including changes in both transcription and trafficking of GLR-1 . In vitro studies using rodent neuron or slice cultures showed that the CaMKIV signaling pathway regulates bidirectional synaptic scaling [18 , 19] . In C . elegans , cmk-1 is the only homolog of mammalian CaMKI and CaMKIV and shares features with both kinases . While the primary sequence of CMK-1 shows more homology to mammalian CaMKI , CMK-1 appears to function more like CaMKIV based on its neuronal expression pattern , its ability to phosphorylate CREB , and its localization to both the cytoplasm and nucleus [23–25 , 51] . Our data show in vivo that the CMK-1/CaMK signaling pathway mediates the feedback mechanism and acts in the nucleus to repress glr-1 transcription ( Figs 3–5 ) . We showed that cmk-1 loss-of-function mutants had increased glr-1 transcript levels based on RT-qPCR and fluorescent reporters ( Fig 3 ) . Analysis of a glr-1 transcriptional reporter in CMK-1 signaling pathway mutants including ckk-/CaMKK1 , cmk-1/CaMK , crh-1/CREB and cbp-1/CBP indicates that the CMK-1 signaling pathway represses glr-1 transcription ( Fig 3 ) . Furthermore , rescue experiments indicate that CMK-1 functions in GLR-1-expressing neurons to repress glr-1 transcription , and this effect is dependent on its kinase activity ( Fig 3H and 3I ) . Several pieces of evidence suggest that in addition to repressing basal glr-1 transcription , CMK-1 also mediates the glr-1 transcriptional feedback mechanism . First , analysis of genetic double mutants between cmk-1 signaling pathway components and glr-1 showed non-additive effects on glr-1 transcription ( Fig 4 ) , consistent with the idea that CMK-1 signaling functions in the same pathway as the feedback mechanism triggered by loss of glr-1 . Second , the feedback mechanism triggered by loss of glr-1 or by expression of constitutively active GLR-1 ( A/T ) regulated the subcellular distribution of CMK-1 between the cytoplasm and nucleus ( Fig 5 ) . These bidirectional changes in GLR-1 signaling had opposite effects on CMK-1 localization to the nucleus , consistent with the idea that decreased GLR-1 signaling results in decreased translocation of CMK-1 to the nucleus whereas increased GLR-1 signaling results in increased translocation of CMK-1 into the nucleus . Third , a gain-of-function allele ( pg58 ) of cmk-1 missing its NES and autoinhibitory domain [27] blocked the glr-1 transcriptional feedback mechanism ( Fig 4C ) . Furthermore , addition of an exogenous NLS to CMK-1 , which forces CMK-1 into the nucleus [28] , was sufficient to inhibit the glr-1 transcriptional feedback pathway ( Fig 5D ) . Together , these data are consistent with a model whereby increased synaptic GLR-1 activates the CMK-1 signaling pathway resulting in increased nuclear accumulation of CMK-1 and repression of glr-1 transcription ( see model in Fig 5E ) . A recent study by Ma et al . , ( 2014 ) using cultured rodent neurons showed that activation of nuclear CaMKIV and phosphorylation of CREB in response to acute stimulation is mediated by the nuclear translocation of γCaMKII [49] . Interestingly , γCaMKII functions in a kinase-independent manner as a shuttle to transport CaM into the nucleus to activate CaMKK and CaMKIV . In contrast , and consistent with previous studies in C . elegans reporting nuclear translocation of CMK-1 in sensory neurons [27 , 28] , our results show that CMK-1 translocates into the nucleus ( Fig 5 ) and regulates glr-1 transcription in a kinase-dependent manner ( Fig 3 ) . Although Ma et al . ( 2014 ) did not investigate the role of γCaMKII in activating CaMKIV in response to chronic changes in activity , our study suggests that mechanisms of activation of nuclear CaMK may differ between mammals and C . elegans . It will be interesting to test whether chronic changes in activity during synaptic scaling in mammalian neurons also require nucleocytoplasmic shuttling of CaM by γCaMKII . Our results suggest that CMK-1 regulates glr-1 transcription both basally and in response to changes in activity . We found that glr-1 transcription increases in cmk-1 signaling pathway mutants ( Fig 3 ) , suggesting that a low level of CMK-1 activity is required to basally repress glr-1 transcription . However , manipulations that increased CMK-1 activity ( i . e . , cmk-1 ( pg58 gf ) mutants ) were not sufficient to repress basal glr-1 transcription , but interestingly , could completely block the increased glr-1 transcription triggered by loss of glr-1 ( Fig 4C ) . This effect of cmk-1 ( pg58 gf ) is reminiscent of a previous finding in which the gain-of-function allele had no effect on basal secretion of neuropeptides from FLP thermosensory neurons but completely blocked heat-induced secretion of neuropeptides [27] . Together , these studies suggest that CMK-1 regulation of basal responses versus activity-induced responses may be differentially controlled . With the exception of a recent report which showed that nuclear Arc represses GluA1 transcription during synaptic scaling [52] , little is known about direct regulation of AMPAR transcription by chronic changes in activity . While several studies have shown that AMPAR mRNA and protein levels are altered during scaling [15 , 53 , 54] and synaptic scaling depends on CaMKIV and transcription [18 , 19 , 21 , 22 , 55] , a direct connection between the CaMK pathway and AMPAR transcription has not been shown . In this study , we showed that bidirectional changes in synaptic activity regulate glr-1 promoter activity in a reciprocal manner and that this effect is mediated by the CaMKK-CaMK signaling pathway . We found that in addition to mutations in ckk-1/CaMKK and cmk-1/CaMK , mutations in crh-1/CREB or cbp-1/CBP also affect glr-1 transcription ( Fig 3 ) . Since mammalian CaMKIV and C . elegans CMK-1 can phosphorylate and activate CREB and CRH-1 , respectively [24 , 25 , 43–46] , these data suggest that the CaMK-CREB axis represses glr-1 transcription . However , this regulation is likely to be indirect because the glr-1 promoter does not contain any canonical CREB binding sites , suggesting that CMK-1 may first activate CRH-1/CREB which in turn regulates transcription of a repressor that controls glr-1 transcription . Interestingly , two recent studies implicate global changes in DNA methylation as a mechanism to regulate AMPAR expression during synaptic scaling [56 , 57] . These studies show that in cultured rodent neurons there is an overall reduction in DNA methylation in response to activity-blockade , whereas DNA methylation increases in response to enhanced neuronal activity . As methylation is typically associated with gene repression , these papers suggest that gene expression increases during synaptic scaling in response to activity-blockade and vice versa . Although the existence of DNA cytosine methylation is controversial in C . elegans , a recent paper showed that adenine methylation and the relevant modifying enzymes do exist in C . elegans and function to regulate transgenerational epigenetic changes [58] . It will be interesting in the future to test if DNA methylation is regulated by CMK-1 signaling to control the glr-1 transcriptional feedback pathway . In conclusion , we identified a novel compensatory feedback mechanism in C . elegans that couples GLR-1 glutamate receptors with their own transcription . We characterized this pathway in vivo and showed that CMK-1 represses glr-1 transcription and translocates between the cytoplasm and nucleus to mediate the feedback mechanism . Regulation of glr-1 transcription in C . elegans and GluR transcription in mammals in response to chronic changes in activity are poorly understood . Future studies are warranted to identify the relevant transcription factors that regulate glr-1 transcription both basally and in response to changes in synaptic activity .
All strains were maintained at 20°C as previously described [59] . The following strains were used for this study: N2 FJ1211 pzEx329 [Pglr-1::NLS-GFP::LacZ::glr-1 3’UTR] FJ1217 pzEx329; glr-1 ( n2461 ) FJ1374 pzEx329; klp-4 ( tm2114 ) FJ1268 pzEx354 [Pnmr-1::NLS-GFP::LacZ::glr-1 3’UTR] FJ1284 pzEx354; glr-1 ( n2461 ) FJ1271 pzEx354; klp-4 ( tm2114 ) FJ1047 pzIs29 [Pglr-1::NLS-GFP::LacZ::unc-54 3’UTR] FJ1109 pzIs29; glr-1 ( n2461 ) FJ1073 pzIs29; klp-4 ( tm2114 ) FJ1375 pzIs29; usp-46 ( ok2322 ) FJ1148 pzIs29; unc-11 ( e47 ) MT6318 eat-4 ( n2474 ) FJ1322 pzEx329; eat-4 ( n2474 ) FJ1237 pzIs29; eat-4 ( n2474 ) FJ1316 pzEx362 [Pglr-1::HisCl1] FJ1352 pzIs29; pzEx362 [Pglr-1::HisCl1] PY1589 cmk-1 ( oy21 ) VC691 ckk-1 ( ok1033 ) FJ1291 pzEx329; ckk-1 ( ok1033 ) FJ1159 pzIs29; ckk-1 ( ok1033 ) FJ1141 pzIs29; cmk-1 ( oy21 ) VC220 cmk-1 ( ok287 ) FJ1376 pzIs29; cmk-1 ( ok287 ) YT17 crh-1 ( tz2 ) FJ1167 pzIs29; crh-1 ( tz2 ) MH2430 cbp-1 ( ku258 ) FJ1288 pzIs29; cbp-1 ( ku258 ) FJ1244 pzIs29; cmk-1 ( oy21 ) ; pzEx333 [Pglr-1::CMK-1] FJ1222 pzIs29; pzEx333 [Pglr-1::CMK-1] FJ1235 pzIs29; cmk-1 ( oy21 ) ; pzEx338 [Pglr-1::CMK-1 ( K52A ) ] FJ1142 pzIs29; glr-1 ( n2461 ) ; cmk-1 ( oy21 ) FJ1214 pzIs29; glr-1 ( n2461 ) ; crh-1 ( tz2 ) GN244 cmk-1 ( pg58 ) ( Gift from Miriam Goodman and Dominique Glauser ) FJ1355 pzIs29; cmk-1 ( pg58 ) FJ1356 pzIs29; glr-1 ( n2461 ) ; cmk-1 ( pg58 ) FJ1310 pzIs29; unc-11 ( e47 ) ; ckk-1 ( ok1033 ) FJ1272 pzEx356 [Pglr-1::CMK-1::GFP] FJ1274 pzEx356; ckk-1 ( ok1033 ) FJ1273 pzEx356; glr-1 ( n2461 ) FJ1364 pzEx356; unc-11 ( e47 ) FJ1354 pzIs29; glr-1 ( n2461 ) ; pzEx370 [Pglr-1::CMK-1::EGL-13 NLS] FJ1246 pzEx342 [Pnmr-1::NLS-GFP::LacZ::unc-54 3’UTR] FJ1377 pzEx342; klp-4 ( tm2114 ) FJ1247 pzEx342; glr-1 ( n2461 ) FJ1347 pzEx329; unc-11 ( e47 ) VM3898 akEx52 [Pnmr-1::GLR-1 ( A/T ) ;lin-15 ( + ) ];lin-15 ( n765ts ) ( Gift from Villu Maricq ) Plasmids were generated using standard recombinant DNA techniques , and transgenic strains were created by plasmid microinjection . Pglr-1::NLS-GFP::LACZ::unc-54 3’UTR was made by cloning 5 . 3 kb upstream of the glr-1 transcription start site into pPD96 . 04 ( Addgene , Fire Lab C . elegans Vector Kit ) containing NLS-GFP::LACZ to generate plasmid FJ#119 and injected at 50 ng/ul to make pzEx260 , which was integrated to make pzIs29 . Pglr-1::NLS-GFP::LACZ::glr-1 3’UTR was made by PCR of the glr-1 3’UTR from CR3 and cloning into pV6 with Sac1 and Spe1 to make pBM7 and then PCR of NLS-GFP::LACZ from Pglr-1::NLS-GFP::LACZ::unc-54 3’UTR and cloning into pBM7 with Nhe1 . pBM12 was injected at 60 ng/ul to make pzEx329 . Pnmr-1::NLS-GFP::LACZ::glr-1 3’UTR was made by digesting Pnmr-1 from pKM05 and swapping into pBM12 for Pglr-1 with Sph1 and BamH1 . pBM17 was injected at 50 ng/ul to make pzEx354 . Pnmr-1::NLS-GFP::LACZ::unc-54 3’UTR was made by digesting Pnmr-1 from pKM05 and swapping into Pglr-1::NLS-GFP::LACZ::unc-54 3’UTR for Pglr-1 with Sph1 and BamH1 . pBM16 was injected at 50 ng/ul with Pmyo-2::mCherry at 10 ng/ul to make pzEx342 . Pglr-1::HisCl1 was made by digesting pNP403 ( Ptag-168::HisCl1::SL2::GFP ) ( Gift from Cori Bargmann ) with Nhe1 and Kpn1 and cloning into pV6 . pBM29 was injected at 5 ng/ul with Pmyo2::mCherry at 10 ng/ul to make pzEx358 . Pglr-1::CMK-1::GFP was made by PCR of CMK-1::GFP from Pttx-1::CMK-1::GFP ( Gift from Piali Sengupta ) and cloning into pV6 with Kpn1 and Sac1 . pBM15 was injected at 2 . 5 ng/ul with Pmyo2::mCherry at 10 ng/ul to make pzEx356 . Pglr-1::CMK-1 was made by PCR of CMK- from Pttx-1::CMK-1::GFP ( Gift from Piali Sengupta ) adding a 3’ stop codon and cloning into pV6 with Nhe1 and Kpn1 . pBM13 was injected at 25 ng/ul with Pmyo2::mCherry at 10 ng/ul to make pzEx334 . Pglr-1::CMK-1 ( K52A ) was made by PCR of CMK-1 ( K52A ) from Pttx-1::CMK-1 ( K52A ) ( Gift from Piali Sengupta ) and cloning into pV6 with Nhe1 and Kpn1 . pBM14 was injected at 25 ng/ul with Pmyo2::mCherry at 10 ng/ul to make pzEx338 . Pglr-1::CMK-1::EGL-13 NLS was made by PCR of CMK-1::EGL-13 NLS from Pttx-1::CMK-1::EGL-13 NLS ( Gift from Piali Sengupta ) and cloning into pV6 with Kpn1 and Sac1 . pBM34 was injected at 25 ng/ul with Pttx-3::GFP at 50 ng/ul to make pzEx370 . L4 wild type and animals expressing Pglr-1::HisCl1 were transferred onto plates with and without 10 mM histamine as previously described [39] . At zero , one , and four hours , animals were picked off plates and the GFP reporter was imaged as described above . Each time point was normalized to the zero hour and then to the corresponding time point of animals no exposed to histamine . Statistics were performed using Student’s t test at each time point comparing animals on histamine to animals off histamine . Locomotion assays were performed as previously described [29 , 60] . Briefly , wild type and animals expressing Pglr-1::HisCl1 were placed on a plate with no food and allowed to acclimate for two minutes . Animals were then transferred to either a plate with or without 10 mM histamine ( no food on either plate ) . Animals placed on histamine plates were exposed for 10 minutes . Animals were then observed for five minutes while reversals were counted manually . Behavioral assays were performed by an experimenter blind to the genotypes being observed . Total RNA was isolated from ten 10 cm plates per genotype of mixed-stage animals by lysing in Trizol ( Invitrogen ) and extracting with an RNeasy Fibrous Tissue Mini kit ( Qiagen ) with on-column DNAse treatment . For each genotype , at least three independent RNA preparations were made alongside a corresponding wild type ( N2 ) preparation to control for variation introduced by each preparation . cDNA from these RNA preps was synthesized using Superscript III Reverse Transcriptase ( Invitrogen ) and oligo d ( T ) primers . RT-qPCR was performed on the MX3000P real-time PCR machine ( Tufts Center for Neuroscience Research ) using the Brilliant SYBR Green QPCR Master Mix . The ΔΔCt method [61] was used to compute the relative amount of glr-1 mRNA compared to two reference genes ( act-1 and ama-1 ) . Primers used for each gene ( all oriented 5’ to 3’ ) : glr-1: F- CCGTTTAAACTTGCATTTGACC , R- ACAGACTGCGTTCACCATGT cmk-1 F- ATGCCCCTTTTTAAGCGACGG , R- ACTGCATACATCTGACCGGCAT act-1 ( DePina , 2011 ) : F-CCAGGAATTGCTGATCGTATGCAGAA , R-TGGAGAGGGAAGCGAGGATAGA ama-1 ( Yan 2009 ) : F- ACTCAGATGACACTCAACAC , R- GAATACAGTCAACGACGGAG SEM was calculated as previously described ( Applied Biosystems ) . Statistical significance was determined using the Student’s t test on the geometric mean of the ΔCt values for each reference gene . | Synaptic homeostasis increases or decreases synaptic strengths in order to stabilize neuronal firing in response to alterations in neuronal activity . Synaptic homeostasis plays an important role during neuronal development and may be deregulated in several neurological diseases . Neurons regulate glutamate neurotransmitter receptor levels at synapses to alter the strength of synaptic signaling during a form of homeostasis termed synaptic scaling . While many molecules have been implicated in synaptic scaling in vitro using cultured rodent neuron or slice preparations , the underlying in vivo mechanisms are not well understood . Here we use the genetic model organism C . elegans to identify in vivo mechanisms involved in a compensatory feedback pathway reminiscent of synaptic homeostasis that couples activity of the glutamate receptor GLR-1 with its own transcription . We show that glr-1 transcription is regulated in a compensatory manner by bidirectional changes in synaptic activity . We find that the CMK-1/CaM kinase signaling pathway represses glr-1 transcription . Furthermore , the subcellular distribution of CMK-1 between the cytoplasm and nucleus is regulated by GLR-1 and is important for mediating the feedback mechanism . This study uses genetics to dissect a negative feedback pathway in vivo and identifies the signaling mechanism that links changes in synaptic activity directly to glr-1 transcription . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"invertebrates",
"neurochemistry",
"chemical",
"compounds",
"mechanisms",
"of",
"signal",
"transduction",
"caenorhabditis",
"gene",
"regulation",
"neuroscience",
"animals",
"organic",
"compounds",
"dna",
"transcription",
"animal",
"models",
"caenorhabditis",
"elegans",
"mod... | 2016 | The CaM Kinase CMK-1 Mediates a Negative Feedback Mechanism Coupling the C. elegans Glutamate Receptor GLR-1 with Its Own Transcription |
Living organisms need to maintain energetic homeostasis . For many species , this implies taking actions with delayed consequences . For example , humans may have to decide between foraging for high-calorie but hard-to-get , and low-calorie but easy-to-get food , under threat of starvation . Homeostatic principles prescribe decisions that maximize the probability of sustaining appropriate energy levels across the entire foraging trajectory . Here , predictions from biological principles contrast with predictions from economic decision-making models based on maximizing the utility of the endpoint outcome of a choice . To empirically arbitrate between the predictions of biological and economic models for individual human decision-making , we devised a virtual foraging task in which players chose repeatedly between two foraging environments , lost energy by the passage of time , and gained energy probabilistically according to the statistics of the environment they chose . Reaching zero energy was framed as starvation . We used the mathematics of random walks to derive endpoint outcome distributions of the choices . This also furnished equivalent lotteries , presented in a purely economic , casino-like frame , in which starvation corresponded to winning nothing . Bayesian model comparison showed that—in both the foraging and the casino frames—participants’ choices depended jointly on the probability of starvation and the expected endpoint value of the outcome , but could not be explained by economic models based on combinations of statistical moments or on rank-dependent utility . This implies that under precisely defined constraints biological principles are better suited to explain human decision-making than economic models based on endpoint utility maximization .
Homeostasis is paramount to all living organisms [1] . Put simply , organisms have to maintain their internal milieu within certain boundaries to avoid dying . This homeostatic principle reverberates on the levels of molecular interactions [2] , hormonal feedback loops [3 , 4] , neural circuits [5] , and psychophysiological processes [6] . Beyond the need for immediate regulation , many species face complex decisions with delayed and probabilistic consequences for long-term metabolic homeostasis . Here , we hypothesize that homeostatic requirements guide foraging decisions in humans . For example , hunting deer provides a large energy gain with low probability of obtaining it , while collecting berries provides a small energy gain with high probability . In order to minimize the probability of starvation , human agents should integrate the statistics of the available options with their current energy levels and with their time horizon . Classical views of homeostasis [1–7] are often illustrated with a thermostat that senses the difference between a temperature set point and the current temperature . This deviation value elicits a change in heating levels . The thermostat is thus supposed to retrospectively compensate deviations that have already manifested themselves . In contrast , we propose that decision-makers can anticipate possible deviations and proactively minimize the probability of reaching a prohibitive boundary such as starvation . This extends established notions of homeostasis in ( psycho ) physiology [1–7] and is concordant with recent theoretical accounts of homeostasis as a principle explaining decision-making in healthy and psychiatric populations [8–10] . When applied to individual decision-making , predictions from this model are in contradistinction to economic models which firmly rest upon axiomatic foundations [11] and elegantly explain many types of monetary decisions [12] . These models posit that decision-makers base their choices on the utility assigned to the endpoint outcome of a choice irrespective of the trajectory to this endpoint [12–14] . In risk-return models and their variants [13 , 15] , the endpoint utility is computed via statistical moments of an outcome distribution , usually expected value , variance , and in some models also skewness . A considerable literature has generalized risk-return models to include subjective transformations of statistical moments [15 , 16] . In behavioral economics , variants of risk-return models have been widely used to describe how humans choose between monetary gambles [17–19] , how they assess real-life events [16] , and how animals decide on primary reinforcers [20 , 21] . Expected utility theory and its derivations constitute another class of models in which values of possible outcomes are transformed into an internal utility measure by a decision-maker's individual utility function [12 , 15 , 22] . Rank-dependent utility models additionally supplement a non-linear weighting of the option’s outcome probabilities [23 , 24] . Similar to risk-return models , rank-dependent utility models have been used extensively to describe empirical data both from the lab and the field [15 , 23 , 24] . Empirically observed deviations from predictions of these microeconomic models are often framed as irrational biases , and additional parameters are included to absorb such biases , but often without making principled assumptions on why these influences should arise in the first place [25] . Here , we furnish a principled biological reason for deviations from economic principles . Critically , maximizing the utility of the endpoint outcome of a given set of options neglects the catastrophic consequences if the trajectory to this outcome reaches a lower bound of the internal milieu . We sought to show that even in a safe laboratory environment , a decision-making model based on homeostatic principles could explain foraging decisions better than economic models . Further , we hypothesized that homeostatic considerations would also guide human decision-making for simply structured lotteries without any reference to foraging , as often employed in behavioral economics [12–14 , 26] . To test these hypotheses , we developed a virtual foraging task . In each trial , human participants chose between two “foraging environments” with different possible “energy” gains and associated probabilities , in which they would forage for up to three consecutive “days” ( see Fig 1A for illustration ) . At the time of choice , an energy bar depicted participants’ current internal state . Participants were instructed that on each foraging day , they would lose one energy point , and gain energy according to the statistics of the chosen environment . Successive days in the task required the integration of risks from multiple foraging attempts . For each trial , participants made a single decision between the two foraging environments for the indicated number of days . Losing all energy points on any day in a given foraging period was framed as “starvation” but was not explicitly punished . Each trial was independent . We did not give feedback on choice outcomes of their choices or intermediate states of the foraging sequence . At the end of the experiment , participants were rewarded for the endpoint foraging outcome of two randomly selected trials . Starvation meant that participants did not win anything from the trial . We hypothesized that participants would compute the probability of starvation for the foraging environments and base their decisions on this metric . We used the mathematics of random walks to analytically derive the distributions of the endpoint outcomes of the foraging period , and of the probabilities of starvation during foraging ( see Fig 1C for a graphical illustration , Fig 1D for an example of the variables derived by this procedure , Table 1 for a summary of the gambles , and Methods and S1 Text for mathematical details ) . Because participants were only rewarded for endpoint outcomes , homeostatic principles are irrelevant for maximizing utility , yet our task was suggestive of using them . Hence , we tested whether such principles also influence decisions when they are not invoked by the task frame . We presented participants with purely economic gambles , framed as wheel-spinning casino lotteries without any reference to foraging ( see Fig 1B ) . These lotteries had identical endpoint outcomes as the foraging environments . The probability of starvation in the foraging frame corresponded to the probability of winning nothing in the casino frame . As in the foraging frame , participants did not receive feedback on the outcomes of their choices . In addition to the instruction , the two frames differed in the fact that options in the foraging frame were presented as gamble sequences when the number of days was greater than one , whereas options in the casino frame were always single-step gambles . To avoid foraging instructions influencing behavior in the casino frame , the casino frame preceded the foraging frame for all participants .
We first asked whether models based on homeostatic principles explain choice better than standard economic models—both in a foraging and in a casino frame . We combined choices from both frames and compared three families of formal decision-making models . The first two model families included variations of two types of economic models while the third family comprised models based on homeostatic considerations ( see Methods and Table 2 for details ) . In line with our hypothesis that participants’ decisions should take into account the probability of starvation ( pstarve ) , the homeostatic model family provided the significantly best fit . Under the assumption that different participants may use different models ( random-effects analysis ) , the exceedance probability that the homeostatic model family is the most frequent in the population was 0 . 9403 ( Table 3 ) . Under the assumption that all participants use the same model ( fixed-effects analysis ) , the winning model belonged to the homeostatic family ( see Fig 2A and Table 2 for log-group Bayes factors based on the Bayesian information criterion ( BIC ) relative to the simplest model; see S2 Text Section 7 and S1 Table and S2 Table for results based on the Akaike information criterion , AIC; see S3 Table for the fits of the different models for each individual participant ) . Thus , the overall comparison of model families confirmed our hypothesis that starvation probability and thus homeostatic principles provided explanatory power in explaining participants’ behavior , over and above economic variables , and although irrelevant for utility maximization in the laboratory . Next , we separately analyzed choices in the foraging frame and in the purely monetary context of the casino frame , by comparing the three model families within each frame . The same overall pattern emerged . The homeostatic family , in which models included pstarve , had the highest exceedance probabilities in both frames independently ( foraging: 0 . 9869; casino: 0 . 8675; Table 3 ) . Also , the models winning in fixed-effects analyses belonged to the homeostatic family ( see Table 2 for log-group Bayes factors based on BIC ) . Further , when we fitted the models separately for the first and second blocks of the foraging and casino frames , the same pattern emerged in all analyses . The homeostatic family had the highest exceedance probabilities and models belonging to this family won the fixed effects analyses ( see Tables 2 and 3; see S2 Text Section 7 , S1 Table and S2 Table for results based on AIC ) . Within the winning model family , we analyzed which specific model best explained choices . In a random-effects analysis , the exceedance probability of the simplest homeostatic model was 0 . 9786 across both frames ( Table 4 ) . Similar results emerged when fitting the models separately within the two frames or separately to the first and second blocks ( Table 4; see S2 Text Section 7 and S4 Table for results based on AIC; see S1 Fig for binned choice data ) . In this winning model , the decision variable was a linear combination of difference in starvation probability ( pstarve ) weighed by a parameter ξ , and difference in expected value ( EV ) . This decision variable was transformed into a decision probability by a sigmoid function with another parameter , β . The previous analyses showed that participants consistently used models based on homeostatic principles in both frames . This leaves open the question how participants used pstarve and whether this differed between the two frames . Hence , we added frame-specific free parameters to the winning model and then tested the parameter estimates across participants . Specifically , we adapted the winning model ( Model 7 ) , which included two free parameters: a parameter β for the decision noise and a parameter ξ to quantify the impact of pstarve on participants’ decision . This parameter ξ was replaced by two frame-specific weighting parameters ( ξforaging and ξcasino ) . We added this model ( Model 10 ) to the initial set of three models in the third family . Despite being penalized for the additional free parameter , it explained choices better than the other three models considered . Its exceedance probability was 0 . 9983 in a random-effects analysis and it had the smallest log-group Bayes factor in a fixed-effects analysis ( Table 5; see S2 Text Section 7 and S5 Table for results based on AIC ) . This indicates that participants weighted pstarve differently in the two frames . Crucially , our prediction that participants’ choices minimize pstarve requires that weighting parameters of pstarve be negative . Indeed , the parameters for the frame-specific parameters ξforaging and ξcasino were significantly smaller than zero across participants ( sign test on parameters in the overall winning model: ξforaging: p< . 001; and ξcasino: p< . 005; Fig 2B ) . That is , participants chose the gambles with the smaller pstarve and thus minimized pstarve . Additionally , across participants ξforaging was smaller than ξcasino ( sign test comparing ξforaging and ξcasino: p< . 05 ) . In line with the above analyses , supporting analyses showed that pstarve played a greater role than EV in the foraging frame , while in the casino frame EV played a greater role than pstarve , for explaining choices ( see S2 Text Section 1 and S6 Table for details ) . Additionally , we devised a supplemental model to test whether the different number of foraging days led to a different weighting of pstarve but did not find evidence supporting this idea ( S2 Text Section 2 and S7 Table ) . We also found no evidence supporting the hypothesis that different combinations of energy levels and foraging days lead to differential weighting of pstarve ( S2 Text Section 4 ) . Supplementary analysis showed that participants did not erroneously include values below zero in their estimation of the outcome distributions in the foraging frame ( S2 Text Section 5 and S8 Table ) . In an exploratory analysis , we also found no evidence for a relationship of the model parameters to participants’ meta-cognitive risk assessments on the domain-specific risk-attitude scale ( S2 Text Section 8 ) . Taken together , participants consistently minimized pstarve in both frames and did so more in the foraging than in the casino frame . Can reaction times ( RTs ) as a tentative measure of choice difficulty give us additional evidence for the relevance of homeostatic principles ? Since our model comparison showed that EV and pstarve explained participants’ choices , we tested whether EV and pstarve also related to RTs . That is , we tested whether RTs were faster for larger absolute differences between the two options in EV and pstarve . This was indeed the case as shown by a linear mixed effects model on log-transformed RTs ( EV: t = -4 . 98 , p< . 001; pstarve: t = -2 . 62 , p< . 05; significance levels were determined by log-likelihood tests , comparing the full model to a model without the respective factor ) . The interaction of EV and pstarve was significant and related to slower RTs ( t = 3 . 52 , p< . 005 ) . See S1 Fig for binned RT data . In sum , a combination of EV and pstarve was related to choice difficulty as indexed by RTs , which corroborates that homeostatic principles guided participants’ choices .
This study addressed whether homeostatic principles explain human decision-making over and above previously described economic models based on endpoint utility maximization . We found that human decisions minimized the probability of reaching a lower homeostatic bound on the trajectory to their endpoint outcomes , despite the fact that our tasks did not entail any explicit negative consequences of reaching this boundary . This was evident both in a virtual foraging task , in which the possibility of starvation was a salient task feature , and in a casino-like frame , in which only the endpoint outcomes of the gambles and their associated probabilities were explicitly stated . Our fine grained model comparison provided evidence that the decision variable in the most parsimonious model was based on a linear combination of the probability of starvation and endpoint expected value ( EV ) , outperforming standard economic models . The maximization of endpoint EV lies at the core of many variants of axiomatically derived microeconomic models . However , neither variants of risk-return models nor variants of expected utility theory predict that decisions minimize the probability of reaching a lower homeostatic bound before that endpoint is realized . The winning model family included the probability of zero outcomes although we did incentivize participants to avoid them , and although zero outcomes are already incorporated into the calculation of statistical moments and utilities . For a description of behavior , we could have used a very specific shape of the utility function which assigns a high negative utility to the zero outcome and positive utility to neighboring positive outcome , in contrast to typical utility functions in the economic literature . However , such a model would neither be more parsimonious than ours , nor offer any additional explanatory power . We note that in the best fitting model , the decision variable was a linear mixture of outcome variables and thus it did not differ from previous risk-return models in its mathematical structure . Crucially , minimization of the probability of a zero outcome provided more explanatory power than risk-attitudes based on variance or skewness . Thus , our results are in line with previous accounts calling for more fine-grained and possibly context-dependent metrics within the framework of risk-return models [19 , 27] . Additionally , our model only makes meaningful predictions when the probability of threats to homeostasis is nonzero and thus our approach has the desirable feature that the scope of the model is under precisely defined constraints . We provide evidence that the homeostatic principle of avoiding a lower boundary on energy levels pervades human decision-making . Classical descriptions often relate homeostatic processes to the actions of a thermostat [6 , 7] . The thermostat example best fits to physiological variables with a narrow homeostatic range , for which this range can be approximated by a set point ( e . g . , blood pH ) [7] . For other variables the homeostatic range is larger . In the case of metabolic homeostasis , glycogen and fat buffers enlarge the homeostatic range and relevant homeostatic counter-measures occur at the boundaries of this range [7] . For simplicity , we assumed starvation to be a hard boundary but the same principle would apply for soft boundaries . More importantly , our results extend the notion exemplified by the thermostat analogy . In line with recent theoretical views on homeostasis in healthy and psychiatric populations [8 , 10] , we conjectured that human decision-makers can estimate the probability of future disruptions to homeostasis . Thus , in contrast to a thermostat that can only react to homeostatic threats once they have occurred , human—and possibly many animal—decision-makers can proactively avoid threats to homeostasis . The same model performed best in both the foraging frame and in the casino frame . We highlight this similarity between the two frames because it shows that the homeostatic principle of minimizing the probability of a zero outcome is at play even when participants are not primed by the task description to do so . Furthermore , the same model explains behavior in gamble sequences and in single-step gambles . In the casino frame , the probability of starvation was directly depicted by the size of the sector in the pie chart that indicated the probability for the zero outcome . Strikingly , in the foraging frame participants integrated the probabilities of gaining energy over the indicated number of days to compute the probability of starvation . Participants could not learn the outcome distributions through experience because we did not provide them with feedback . Thus , decisions in the foraging frame were not dependent on participants having directly experienced sequences in the virtual foraging environments . Risky decision-making differs depending on whether outcome distributions are described or learned from experience [22 , 28] . For example , rare events tend to exert less impact in decisions based on experience . Our results suggest that such an underweighting might not occur for the probability of starvation [28] . Within the winning model , more fine grained analyses revealed differences between the two frames in the best-fitting parameter estimates . The probability of starvation in the foraging frame had a greater impact on participants’ decisions than the corresponding probability of receiving nothing in the casino frame—an effect unrelated to the sequential versus single-step presentation of the gambles ( see S2 Text Section 3 ) . This was the case even though participants had to compute the probability of starvation in the foraging frame by combining information about internal state , foraging options , and time horizon . Approximating starvation probabilities may become more difficult and thus imprecise as the number of steps increases . In the current study , participants were able to approximate the probability of starvation with sufficient accuracy for at least three steps , as their decisions were based on this metric . The structure of our tasks complies with the requirements of economic paradigms such as complete knowledge and incentive-compatibility [12] . Thus , specific task characteristics are unlikely to explain why our homeostatic model outperformed standard economic models , based on statistical moments [15 , 17 , 18] or non-linear probability weighting [23 , 24] . Instead , we reason that the biological constraints relevant in ecological contexts such as hunting or farming exert a prevailing impact on human decisions in the laboratory—even if apparently irrelevant to the task at hand . A similar rationale has recently been advocated in discussions of whether animal [29 , 30] and human [31–34] decision-making deviates from normative models . According to probabilistic accounts of brain function , the brain uses prior probabilities to perform probabilistic inferences [8 , 14 , 35 , 36] . These prior probabilities are tuned—by evolution and/or experience—to the natural statistics of real world environments [30 , 31] . Consequently , human and non-human decision-makers may behave rationally according to their beliefs but they appear irrational because those beliefs are not warranted in deliberately simplified laboratory tasks or in some other contexts [30 , 37] . Overall , this recent approach argues for complementing considerations about economic rationality ( i . e . , maximizing financial gain given currently available information ) with considerations about ecological rationality ( i . e . , maximizing fitness given priors on environmental statistics ) . Its promise lies in unifying and explaining a diverse set of seemingly irrational behaviors while its challenge lies in identifying and testing the ecological principles on which to base such explanations [30] . The current study demonstrates that a basic biological principle about the internal milieu provides a refined and parsimonious explanation of human decisions under risk . Our virtual foraging task was specifically designed to test the influence of homeostatic principles on risky choice . It thereby relates to , and extends , previous tests of risk-sensitive foraging theory in animals [38–41] . Risk-sensitive foraging theory provides an account of how animal should choose between risky foraging options so as to maximize their fitness [40] . The crucial insight of risk-sensitive foraging theory is that foraging animals should choose options with higher variance if options with lower variance cannot provide a sufficient amount of energy to meet critical levels until a certain time point . For example , hungry birds in winter should become more risk-prone as nightfall approaches . Thus , risk-sensitive foraging theory provides an ecologically rational benchmark [38–40] although empirical evidence for it has been mixed [40] . Similar to risk-sensitive foraging theory , our model comprises a hard boundary that is relevant to the decision-maker within a given time horizon . Crucially , we introduce a novel and simple mathematical description for deriving sequential gambles that mirror foraging settings . Testing the model in a virtual setting in humans circumvents challenges of non-human animal research such as the need to impose actual threats onto participants or the need to impart outcome distributions through extensive training . Risk-sensitive foraging theory has been related to loss aversion [40] , which refers the empirical observation that humans seem to care more about losses than gains of equivalent magnitude [40 , 42] . One may speculate that loss aversion may be related to our finding that participants minimized the probability of starvation . However , loss aversion can only arise in mixed gambles ( i . e . , when options entail gains and losses ) [23 , 42] , and our gambles did not involve losses . Therefore , loss aversion cannot explain our findings . Our approach is in line with some recent studies that have employed virtual foraging-like tasks to probe the psychological and neural mechanisms of complex decision-making in animals [41 , 43] and humans [44–46] . One notable study showed that humans adjust their risk-taking behavior dynamically over a sequence of gambles [44] . Another study provides evidence that humans continuously reassess the sequences of gambles available to them in the future although economically optimal strategies prescribe that decisions be independent of sequence order [47] . Our results complement these findings by suggesting that such behavior could be easily explained if people take into account the probability of “starvation” during the choice sequence . Overall , the current study makes detailed predictions for apparent irrationalities in dynamic foraging tasks that are consistent with earlier reports . Our model of homeostatic decision-making lends itself to possible extensions . First , decision-makers usually have to maintain several variables in a homeostatic range . Our model can easily be extended to such situations with the prediction that decision-makers minimize the joint probability of starvation , which may imply giving up a large amount of one variable to avoid getting zero of another . When boundaries are soft rather than hard , this can be thought of as minimizing a constrained functional that describes a trajectory through homeostatic space . Second , risk preferences are often assumed to be rather stable personality traits [48] but our model implies that they should vary depending on threats to homeostasis [38] . Third , insurances for rare high-impact events have been a recent focus in economics [49] . The concept of starvation in our model may give a handle on investigating the impact of such events on human decisions . Our results suggest that the pursuit of this fundamental biological goal translates into simple but specific predictions for decision-making that are amenable to empirical tests . Standard economic models provide an indispensable benchmark against which to test the inclusion of additional considerations about biological considerations [12 , 15 , 23] . Commonly , models of risky decision-making have to strike a balance between the elegance of axiomatic economic foundations that are at odds with empirical observations and the unwieldy ad hoc assumptions of irrational biases . Our results provide an example that models based on fundamental biological principles such as homeostasis can reconcile parsimony with an explanation for apparent irrationalities .
The study was conducted in accord with the Declaration of Helsinki and approved by the governmental research ethics committee ( Kantonale Ethikkommission Zürich , KEK-ZH-Nr . 2013–0328 ) . All participants gave written informed consent using a form approved by the ethics committee . Twenty-two participants ( 15 female; age: mean = 25 years , SD = 5 . 0 ) were recruited from a student population via mailing lists of local universities . Participants were paid a show-up fee of CHF 15 plus a variable amount ( see below ) . Participants completed 960 trials in two variants ( frames ) of a binary choice task: the foraging and the casino frames ( Fig 1A and 1B ) . The same list of 480 combinations of gambles was used for both frames ( i . e . , the outcome distributions were numerically equal; see Table 1 for an overview of the variables; see below and S1 Text for details on how gambles were derived ) . For both frames , participants received detailed written instructions and performed eight training trials followed by two blocks of the actual task . The task was presented using the MATLAB toolbox Cogent 2000 ( www . vislab . ucl . ac . uk ) . The instruction for the foraging frame told participants to imagine themselves in a hunter-gatherer context . Since we wanted to exclude that putting participants into a foraging mindset influenced choices in the casino frame , all participants completed the casino frame before the foraging frame . The 480 gamble combinations for each frame were split into two blocks , which were counterbalanced for order . Each list contained 80 unique gamble combinations; the remaining 400 gamble combinations were included in both lists . During the game participants did not see the outcomes of their choices . That is , participants were given examples of possible outcomes in the written instruction but they did not directly experience them . At the end of the experiment , one trial from each of the four blocks was randomly chosen . The outcomes of these trials were determined based on participants’ choices and the corresponding amount was paid out ( 1 point was worth CHF 0 . 75 ) . Thus , both frames were incentivized in the same way . See Fig 1 and S2 Text for further details . We used the mathematics of random walks to derive outcome distributions for the 480 combinations of gambles ( Table 1 ) . We briefly introduce the basic logic ( Fig 1C ) . For details see S1 Text . In a random walk an imaginary agent starts at a given position on a line of positive integers . The starting position corresponds to the initial number of energy points . The agent makes a number of steps on that line , which correspond to the number of days . In each step , the agent moves “right” with a certain probability p and “left” with the probability q = 1-p . Moving left corresponds to unsuccessful foraging and the step sizes correspond to a fixed cost of one energy point . Moving right corresponds to successful foraging and the step sizes correspond to the variable points to gain ( minus the cost of one point ) . Zero represents an absorbing boundary ( i . e . , if the agent reaches zero , the random walk stops ) . The possible positions on the number line after a certain number of steps correspond to the range of outcomes . To obtain the probability of an outcome , all the probabilities of all “branches on the tree” toward that outcome have to be summed up . ( The number of “branches” is calculated with a binomial coefficient . ) Along a given branch the probabilities ( i . e . , p or q ) have to be multiplied . We created different gambles by varying combinations of starting positions , probabilities of moving right and step sizes to the right . In the current study , we included gambles with four different combinations of starting positions and number of steps ( a ) starting position 1 and 1 step , ( b ) starting position 1 and 2 steps , ( c ) starting position 2 and 2 steps , and ( d ) starting position 2 and 3 steps ( with each combination occurring in 120 combinations of gambles ) . We used the outcomes and their respective probabilities to calculate the statistical moments of the chosen gambles . The probabilities of reaching zero are denoted pstarve . Note that in the gambles included in the current study pstarve was never zero ( see Table 1 ) . Log-transformed RTs were analyzed using a linear mixed effects model as implemented in the R package lmer [53] ( http://cran . r-project . org/web/packages/lme4/index . html ) . Log-transformed RTs were approximately normally distributed . The independent variables in the mixed effects model were the variables that the comparison of choice models identified as relevant ( i . e . , EV and pstarve ) . Specifically , the fixed effects of the model included the difference between the two options in EV and pstarve as well as the interaction of the two . Random effects for participants included a random intercept and random slopes for EV , pstarve , and their interaction . The model is given by the following equation: Significance levels of the fixed effects were determined by performing log-likelihood tests , which compared the full model to models without the respective factor . | Common decision-making models arise from firm axiomatic foundations but do not account for a variety of empirically observed choice patterns such as risk attitudes in the face of high-impact events . Here , we argue that one reason for this mismatch between theory and data lies in the neglect of basic biological principles such as metabolic homeostasis . We use Bayesian model comparison to show that models based on homeostatic considerations explain human decisions better than classic economic models—both in a novel virtual foraging task and in standard economic gambles . Specifically , we show that in line with the principle of homeostasis human choice minimizes the probability of reaching a lower bound . Our results highlight that predictions from biological principles provide simple , testable , and ecologically rational explanations for apparent biases in decision-making . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Maintaining Homeostasis by Decision-Making |
Rhodnius nasutus , a vector of the etiological agent Trypanosoma cruzi , is one of the epidemiologically most relevant triatomine species of the Brazilian Caatinga , where it often colonizes rural peridomestic structures such as chicken coops and occasionally invades houses . Historical colonization and determination of its genetic diversity and population structure may provide new information towards the improvement of vector control in the region . In this paper we present thoughtful analyses considering the phylogeography and demographic history of R . nasutus in the Caatinga . A total of 157 R . nasutus specimens were collected from Copernicia prunifera palm trees in eight geographic localities within the Brazilian Caatinga biome , sequenced for 595-bp fragment of the mitochondrial cytochrome b gene ( cyt b ) and genotyped for eight microsatellite loci . Sixteen haplotypes were detected in the cyt b sequences , two of which were shared among different localities . Molecular diversity indices exhibited low diversity levels and a haplotype network revealed low divergence among R . nasutus sequences , with two central haplotypes shared by five of the eight populations analyzed . The demographic model that better represented R . nasutus population dynamics was the exponential growth model . Results of the microsatellite data analyses indicated that the entire population is comprised of four highly differentiated groups , with no obvious contemporary geographic barriers that could explain the population substructure detected . A complex pattern of migration was observed , in which a western Caatinga population seems to be the source of emigrants to the eastern populations . R . nasutus that inhabit C . prunifera palms do not comprise a species complex . The species went through a population expansion at 12–10 ka , during the Holocene , which coincides with end of the largest dry season in South America . It colonized the Caatinga in a process that occurred from west to east in the region . R . nasutus is presently facing an important ecological impact caused by the continuous deforestation of C . prunifera palms in northeast Brazil . We hypothesize that this ecological disturbance might contribute to an increase in the events of invasion and colonization of human habitations .
Chagas disease is caused by the protozoan Trypanosoma cruzi ( Kinetoplastida , Trypanosomatidae ) and transmitted primarily through the feces of infected triatomine bugs ( Hemiptera , Reduviidae ) [1] . Endemic to Latin America and the Caribbean , it is estimated that approximately 6–7 million people are infected worldwide [2] . Although curable when treated early with antiparasitics in the acute phase , there is no vaccine available and treatment of chronic patients only reduces serum parasite detection , but not cardiac clinical complications [3] . Therefore , control programs have placed their efforts on the elimination of domestic vectors , by spraying insecticides indoors [4] . A recurrent problem seems to be the constant re-infestation of insecticide-treated households by abundant native triatomine species , which consists of a major challenge for the consolidation of Chagas disease control efforts [5 , 6] . This dreadful scenario is often reported in Brazil where native T . cruzi-infected triatomines continuously invade houses in both rural and urban areas maintaining the risk of transmission [7 , 8] . The Caatinga and the Cerrado biomes together , harbor most of the triatomine species diversity in Brazil [9] . Of the 68 triatomine species known to occur in the country , 10 have successfully adapted to the harsh droughts imposed by the Caatinga . Rhodnius nasutus Stål , 1859 is one of these species , frequently reported in bird and mammal nests on the crown of Copernicia prunifera ( Carnaúba ) palm trees in the vicinity of human habitations . It is often naturally infected with the T . cruzi parasite and is capable of invading rural and urban houses and colonizing peridomestic structures [10 , 11] . It can also occur in Attalea speciosa , Mauritia flexuosa , Syagrus oleracea and Acrocomia intumescens palm trees [12 , 13] . Interestingly , R . nasutus has also been reported to occur in Licania rigida ( Oiticica ) dicotyledon trees typical of northeast Brazil [10 , 14] . Determination of R . nasutus population structure in the Caatinga biome and estimation of present gene flow may contribute to a better understanding of the species' historical colonization process as well as provide new information towards the improvement of vector control in the region . In this paper we present the phylogeography and demographic history of this Chagas disease vector in the Caatinga inferred with mtDNA ( cyt b ) and nuclear ( microsatellites ) markers .
Two-hundred and twenty-eight Rhodnius specimens were collected after the inspection of 53 Copernicia prunifera palms in eight geographic localities within the Brazilian Caatinga biome ( Table 1 ) . The two closest collection sites are located 40 km apart ( Altos and Campo Maior ) , while the most distant are 670 km apart ( Parnaíba and Serra Talhada ) . The Caatinga biome covers nearly 85 , 000 km2 of the northeast region , belongs to the seasonally dry tropical forests phytogeographic unit [15] , and is a species-rich xeric environment [16 , 17] . Insect field captures were approved by the Instituto Chico Mendes de Conservação da Biodiversidade ( ICMBio ) and were carried out following Gurgel-Gonçalves et al . [18] . Briefly , after reaching up to the palm tree crowns with a ladder , leaves and fibers were pruned and placed into plastic bags . The collected material was then lowered to the ground and scattered over a piece of white cloth to facilitate bug detection . Rhodnius nasutus specimens were identified morphologically based on Lent and Wygodzinsky [19] and confirmed by molecular taxonomy based on a 682 base pair ( bp ) fragment of the mitochondrial cytochrome b gene ( cyt b ) [20] . Sequences generated with the amplification of this fragment were also used in the population genetics and phylogeographic analyses described in the next section . DNA extraction was performed using the Promega Wizard Genomic DNA extraction kit ( Promega , Madison , Wisconsin , USA ) following the manufacturers recommendations . A 682-bp fragment of the cyt b gene was PCR-amplified as described in Monteiro et al . [20] . Amplicons were purified with the Hi Yield Gel/PCR DNA extraction kit ( Real Biotech , Banqiao , Taipei , Taiwan ) , or following the purification protocol with PEG-NaCl 20% - 5mM ( modified from [21] ) . Both DNA strands were subjected to Sanger sequencing reactions with the ABI Prism BigDye Terminator v3 . 1 Cycle Sequencing kit ( Thermo Fisher Scientific , Walthan , Massachusetts , USA ) and run on an ABI 3730 automated sequencer ( Applied Biosystems , Foster City , California , USA ) . The removal of primer sequences , editing of both forward and reverse strands , and the generation of a consensus sequence for each sample were performed with Seqman Lasergene v . 7 . 0 ( DNAStar , Inc . , Madison , Wisconsin , USA ) . Consensus sequences were aligned to published ortholog sequences from other members of the ‘robustus lineage’ ( Rhodnius robustus I-V , R . prolixus , R . neglectus , R . nasutus and R . barretti ) and analyzed to confirm species identity based on genetic distances using Mega v . 5 [22] . Taxonomic identification was achieved based on the genetic similarity with other Rhodnius sequences available in the GenBank public database through the Basic Local Alignment Search Tool ( https://blast . ncbi . nlm . nih . gov/Blast . cgi ) , considering the following parameters: percentage of identity and coverage , e-value , and also Bayesian phylogenetic analyses . A Bayesian phylogenetic cyt b tree based on the 147 samples of morphologically-identified R . nasutus and the other members of the ‘robustus lineage’ was inferred in BEAST v . 1 . 8 [23] . Rhodnius barretti [24] was used to root the maximum clade credibility tree . The best fit model of substitution was determined using jModeltest v . 2 [25] . We tested whether a strict or a relaxed molecular clock best fitted our data through a Bayesian random local clock analysis ( RLC ) [26] . Three independent runs were performed for 109 generations , sampling every 20 , 000 generations . Convergence of parameters and proper mixing were confirmed through the calculation of effective sample sizes ( ESS ) in Tracer v . 1 . 6 [27] , and ESS estimates above 104 were considered appropriate [28] . Gene tree and species tree were reconstructed with *BEAST [29] . Information in the literature about the taxonomic identification of DNA sequences retrieved from GenBank was used to assign prior information on each sequence to a particular species . The 147 cyt b sequences from field-collected specimens without molecular identification were labeled as “Rhodnius sp . ” . Suggested divergence rate for triatomines of 1 . 1 to 1 . 8% per Myr was used as a prior with normal distribution [30] . Molecular indices of haplotype diversity ( Hd ) and nucleotide diversity ( π ) were computed in DnaSP v . 5 [31] for each population and sequence divergence between populations were calculated in Mega v . 5 [22] . A median-joining network [32] was constructed with Network v . 4 . 6 ( Fluxus Technology Ltd . 2008 ) for a better visualization of the relationships between R . nasutus cyt b haplotypes . Since samples from Jaguaruana were collected in 2004 and all other samples in 2010 , we tested whether this could have an impact in our study by carrying out an analysis of variance ( AMOVA ) grouping samples according to the collection date with Arlequin v . 3 . 5 [33] . Levels of genetic differentiation among populations were determined with Wright’s pairwise FST comparisons [34] in Arlequin v . 3 . 5 [33] . Correction for multiple testing was done with the false discovery rate ( FDR ) method [35] under 5% of significance level ( P-value = 0 . 013 ) . The coefficient of determination R2 to evaluate the fit of data to theoretical models of distribution of genetic distances under different demographic scenarios was calculated in the R environment [36] . Deviations from neutrality were assessed with Fu’s Fs [37] and Tajima’s D [38] tests in DnaSP v . 5 [31] . Significant and negative values of both tests indicate population size expansion or purifying selection . Mismatch distribution tests were also used to infer possible demographic and spatial expansions [39] . We also used Beast v . 2 . 4 [40] to reconstruct a Bayesian skyline plot ( BSP ) and test which of the two different demographic models ( constant population size or exponential growth ) better explained the demographic history of R . nasutus . The best model was selected through the comparison of marginal likelihood , calculated with the path sampling algorithm [41] under 106 MCMC chain length and 60 steps , based on Bayes Factors ( BF ) [42] . In all cases , results from two independent runs ( 2 x 107 generations with the first 2 x 106 discarded as burn-in and parameter values sampled every 2 x 103 generations ) were analyzed with Tracer v . 1 . 6 [27] . Convergence of parameters and proper mixing were confirmed through the calculation of ESS , and estimates above 104 were considered appropriate [28] . Because only a single species was analyzed ( and thus it is highly probable that specimens from all populations evolve at a single rate ) , we imposed a strict clock in the analysis . Suggested divergence rates for triatomines of 1 . 1 to 1 . 8% per Myr were used [30] . We matched the R . nasutus demographic history to paleoclimatic events estimated through oxygen isotope marine paleotemperature records [43] , used as etalon for the study of global climate changes . Microsatellite alleles of eight loci were amplified in a total reaction volume of 10 uL containing 1 unit of TaqGold DNA polymerase , 2 mM of each dNTP , 1 . 5 mM MgCl2 , 10 mM of magnesium free Buffer 10X ( Thermo Fisher Scientific Co . , Walthan , Massachusetts , USA ) , 5 pmol of each primer and approximately 10 ng of extracted DNA . PCR conditions were as follows: 1 min at 95°C , 30 cycles of 30 s at 94°C , 30 s at Ta °C and 45 s at 72°C , and 72°C for 30 min; Ta is the annealing temperature for each locus ( S1 Table ) . All PCR products were run with a size standard GS500 LIZ on an ABI 3730xl Genetic Analyzer , and allele fragment lengths quantified using the Peak Scanner software v . 1 . 0 ( Applied Biosystems , Foster City , California , USA ) . Because these primers are heterologous ( designed to amplify microsatellite loci of other Rhodnius species ) [44–46] , the orthology of microsatellite regions and repetition motifs were molecularly validated . Amplified loci were submitted to Sanger sequencing reactions with the ABI Prism BigDye Terminator v3 . 1 Cycle Sequencing kit ( Thermo Fisher Scientific , Walthan , Massachusetts , USA ) and run on an ABI 3730 automated sequencer ( Applied Biosystems , Foster City , California , USA ) . Microsatellite genotypes were screened for likely scoring errors , allele dropout , and presence of null alleles with Micro-Checker v . 2 . 2 [47] . Number of shared genotypes , number of private alleles for each sampling site and Shannon’s allele information index ( SHA ) as a measure of gene diversity were performed with GenAlEx v . 6 . 5 [48] . Deviations from Hardy–Weinberg equilibrium and tests of linkage disequilibrium for each locus were performed with Arlequin v . 3 . 5 software [33] . Genetic differentiation between populations ( pairwise FST ) between sampling localities and inbreeding coefficients ( FIS ) were carried out with Arlequin v . 3 . 5 software [33] . Sampling localities for which FST was non-significant were considered as belonging to the same population . Mutual index ( SHUA ) [49] were also estimated with GenAlEx v . 6 . 5 [48] . The Bayesian clustering program Structure v . 2 . 3 [43] was used to estimate population assignment without prior assumptions of population subdivision . We used the admixture model due to the lack of information about the ancestry of the field-collected specimens , with correlated allele frequencies , which means that these frequencies in different populations are likely to be similar as a consequence of migration or shared ancestry [50] . Burn-in and simulation were set at 2 . 5 x 105 iterations and 106 Markov Chain Monte Carlo ( MCMC ) generations , respectively . Ten independent runs were performed for each value of K ( for 2–8 ) , as suggested by Pritchard et al . [50] . The most likely value of K was estimated with the ΔK method [51] . The Ewens–Watterson neutrality test [52] , which relies on the comparison of observed and expected homozygosity , was performed with the Arlequin v . 3 . 5 software [33] . The program Bottleneck v . 1 . 2 [53] was used to test bottleneck events by evaluating deviations of mutation-drift equilibrium . Three different mutation models of microsatellite evolution were employed: Infinite Allele model ( IAM ) , Stepwise Mutation Model ( SMM ) , and Two-Phased model ( TPM ) , which is an intermediate to the SMM and IAM as it incorporates the mutational process of the former , but allows for mutations of a larger magnitude to occur . The significance was assessed with the Wilcoxon sign-rank test [54] , a more powerful and robust test when few polymorphic loci are available ( < 20 ) . We also constructed a population network with EDENetwork v . 2 . 1 [55] to determine possible source and sink populations of R . nasutus . Basically , this analysis reconstructs a minimum-spanning tree based on the pairwise FST matrix and calculates three different parameters: ( a ) degree , which is defined as the number of connections a node has in the network , summarizing which populations are exchanging migrants; ( b ) clustering , which measures the number of subpopulations within a population; and ( c ) betweenness , which is the number of shortest paths between populations , reflecting areas of intense gene flow . These measures allow identifying “source” and “sink” populations ( i . e . those with the highest degree and betweenness ) [56] . To investigate the patterns of gene flow between the populations defined by Structure we performed a Bayesian approach implemented in Migrate-n v . 3 . 2 [57] . Parameters were first estimated under a full migration model that allowed gene flow to occur among all regions . Then we tested 10 reduced models ( based on FST results ) representing different patterns of migration . Migrate-n analysis was performed with a single long cold chain and four hot swapping chains using a static heating scheme ( temperature: 1 . 0 , 1 . 5 , 3 . 0 , 106 ) , with two independent runs , sampling at every 100th step for a total of 2 x 105 recorded steps ( burn-in = 3 x 104 steps ) . The 10 migration models were compared based on their marginal likelihood and probability using thermodynamic integration with Bezier approximation ( implemented in Migrate-n ) according to Beerli and Palczewski [58] . The plot for visualizing of migration pattern inferred by Migrate-n analysis was drawn in the R environment [36] using Migest Package [59] . The migration scenario with highest probability was tested through an AMOVA with Arlequin v . 3 . 5 [33] . Isolation by distance was estimated for cyt b and microsatellite data through linear regression analyses of dependence , comparing the logarithms of genetic distances ( pairwise FST values ) [60] , and geographic distances between the eight localities , in the R environment [36] . The significance level of hypothesis testing was set at α = 0 . 05 .
Thirty-three of the 53 C . prunifera palm trees investigated ( 62% ) were infested with Rhodnius specimens . Higher infestation percentages were detected in deforested areas ( Table 1 ) . Thirty-three palm trees were sampled in four forested locations and showed lower infestation percentages ( Mean: 58% , range = 50–80% ) than the 15 palm trees collected in four deforested areas ( Mean: 87% , 75–100% ) . From the 228 Rhodnius specimens collected , 163 insects were morphologically identified as Rhodnius nasutus . Only two of the DNAs extracted from all 157 specimens resulted in unsuccessful amplification of the desired cyt b fragment . Eight specimens exhibited low-quality sequences ( i . e . < 500-bp ) and thus were discarded from the analysis . Therefore , our dataset included 147 cyt b sequences of 595-bp long from R . nasutus individuals sampled from eight localities in the Brazilian Caatinga ( Table 1 ) . Pairwise comparison of the sequences obtained with those available in GenBank revealed 99–100% identity with another R . nasutus “reference” sequence ( JX273155 ) generated by our group [24] . This sequence was obtained from a specimen collected in Jaguaruana , Ceará , where the morphologically similar species R . neglectus is not known to occur [61] . Hasegawa-Kishino-Yano ( HKY [62] ) was selected as the best evolutionary model for the data , following the Akaike and Bayesian Information criteria . RLC analysis showed that a “strict” clock is more suitable to our cyt b dataset ( rate change: median = 0; variance = 0 . 67; ESS = 4020 ) . Bayesian species tree ( Fig 1 ) revealed that all sequences from Rhodnius specimens collected formed a monophyletic and well-supported clade with the R . nasutus reference sequence ( PP = 1 . 0 ) , corroborating further their taxonomic identity . A bayesian coalescent gene tree disclosed the short branches ( i . e . low sequence divergence ) among R . nasutus samples . Although posterior probabilities for R . nasutus clades were low ( PP < 0 . 9 ) , it is noticeable that only the R . nasutus sequences from Carnaúba dos Dantas and Sousa clustered in separate clades without the presence of at least one sequence from another locality . Sequences from the other six localities clustered together in other three different clades . AMOVA analysis between Jaguaruana and the other populations did not indicate any sampling effect of collecting at different times as genetic differences between these groups were not significant ( ΦST = 0 . 18 , P = 0 . 98 ) . Therefore , these samples were included in further analyses . Molecular divergence of R . nasutus cyt b sequences varied between 0%-0 . 8% ( S2 Table ) , as expected for intraspecific comparisons in Rhodnius species [20] . Pairwise comparisons within and between localities ranged from 0% to 0 . 2% and 0 . 2% to 0 . 8% , respectively . Overall , the most divergent sequences were present in Parnaíba ( mean divergence = 0 . 62% ) and the less divergent sequences were in Jaguaruana ( mean divergence = 0 . 27% ) . Inspection of the sequences revealed 15 polymorphic sites and 16 haplotypes , with two haplotypes ( 3 and 9 ) shared between different localities ( Fig 2A ) . Molecular diversity indices ( Table 2 ) showed high haplotype diversity ( Hd = 0 . 45–0 . 83 ± 0 . 10 ) , but low nucleotide diversity between haplotypes ( π = 0 . 0032 ± 0 . 0002 ) . The highest haplotype diversities were found in Piracuruca ( N = 9 , Hd = 0 . 694±0 . 147 ) and Altos ( N = 4 , Hd = 0 . 667±0 . 204 ) , and the lowest haplotype diversity in Sousa ( N = 28 , Hd = 0 . 071±0 . 065 ) . All sequences from Carnaúba dos Dantas ( N = 25 ) were identical . The highest nucleotide diversities were found in sequences from Parnaíba ( N = 18 , π = 0 . 00253±0 . 00073 ) and Piracuruca ( N = 9 , π = 0 . 00205±0 . 00058 ) , and the lowest nucleotide diversities in sequences from Sousa ( N = 28 , π = 0 . 00012±0 . 00011 ) and Jaguaruana ( N = 32 , π = 0 . 00021±0 . 00013 ) . Pairwise population FST estimates are given in Table 3 . Most of the comparisons revealed high and significant values ( > 0 . 6 ) . Non-significant FST values were observed only when the populations of Altos , Campo Maior and Jaguaruana were compared . The excessive number of high FST values is due to the presence of 14 ( out of 16 ) unique haplotypes . Haplotypes derived from cyt b sequences revealed a network ( Fig 2B ) with two central haplotypes ( haplotypes 3 and 9 ) that are very abundant and widespread , and to which several less common haplotypes are closely related ( 1–2 mutational steps ) . The most frequent haplotype ( Haplotype 9 , N = 46 ) is shared with specimens from Altos , Campo Maior and Jaguaruana , and the other haplotype ( Haplotype 3 , N = 9 ) is shared with specimens from Campo Maior , Jaguaruana , Parnaíba and Serra Talhada . This type of haplotype connection suggests population expansion or retention of ancestral polymorphism with little migration , evidenced by the geographically restricted haplotypes . Considering that all R . nasutus sequences are separated by small genetic distances ( most haplotypes are connected by a single mutational step ) , neutrality tests , mismatch distribution and BSP analyses were performed with the entire dataset . Mismatch distribution analysis ( Fig 3A ) exhibits a relatively good fit to the expected mismatch distributions under the model of population expansion ( coefficient of determination R2 for population growth model is 0 . 95 ( p = 0 . 008 ) while for constant population size model is 0 . 72 ( p = 0 . 160 ) ) . Neutrality tests did not indicate significant departures from neutrality ( Fu’s Fs = -4 . 27 , p = 0 . 085; Tajima’s D = -0 . 75 , p = 0 . 246 ) . Bayesian analyses of population growth revealed that marginal likelihood for constant population was -1240 . 82 and for exponential growth was -1228 . 27 . Thus , the demographic model with the best fit was the exponential growth model ( BF = 25 . 11 ) . Demographic history of the species ( Fig 3B ) depicts a population expansion that started at about 10 thousand years ago ( ka ) , which coincided with a global climate warming corresponding to the first Marine Isotope Stage ( MIS ) . Migrate-n analysis with the cyt b dataset resulted in no clear pattern of migration probably due to the small amount of variation among sequences , and thus it was not included in the paper . Eight microsatellite loci ( S1 Table ) were analyzed for 155 of the 157 R . nasutus specimens collected in the Caatinga biome . The number of alleles per locus ranged from 1 ( R8 ) to 14 ( List14-064 ) . Since the R8 locus was monomorphic for all individuals , it was excluded from the analysis . The basic statistics for the microsatellite data is summarized in Table 4 . Three loci showed significant deviations from HWE in at least one population , possibly due to the admixture of null alleles or LD . Microsatellite loci analyses revealed moderate levels of genetic diversity and low allelic diversity . Null alleles were detected for loci List14-010 in Group 1 and List14-021 in Group 2 . Significant LD between List-025 and List-064 loci , and List-025 and L43 were found in Group 1 . Among 155 individuals , 111 had unique genotypes: ALT+CAM– 22; PAR– 18; PIR– 8; CAR– 17; JAG– 25; STA– 9; SOU—15 . There are two shared genotypes between CAR and JAG , one genotype shared between STA and JAG , one shared between PAR and PIR , and another shared between PAR and SOU . The highest number of shared genotypes was observed among R . nasutus specimens from Jaguaruana . In addition , positive FIS values were significant in Group 2 , which comprised Jaguaruana specimens . The Bayesian clustering using seven microsatellite loci for 155 individuals revealed the existence of four groups ( Fig 4 ) . The ΔK method indicated 4 groups as the most likely population structure . Group 1 includes all sampling locations of R . nasutus in Altos ( ALT ) , Campo Maior ( CAM ) , Parnaíba ( PAR ) , and Piracuruca ( PIR ) . Group 2 is the most numerous and genetically variable and contains samples from Jaguaruana ( JAG ) and Carnaúba dos Dantas ( CAR ) . Group 3 includes specimens from Serra Talhada ( STA ) , and Group 4 is composed of specimens from Sousa ( SOU ) . These groups were in accordance with the pairwise FST comparisons among sequences from different localities . Pairwise FST values between sampling localities ( Table 3 ) revealed non-significant differences between Altos , Campo Maior , Piracuruca and Parnaíba , and also between Jaguaruana and Carnaúba dos Dantas . Pairwise FST values between Sousa , Serra Talhada and all other localities were significant . The highest value of Shannon’s allele information index ( SHA ) describing genetic variation in population was shown in Group 1 ( 0 . 955 ) . Group 1 also displayed the highest number of private alleles ( 10 ) over all loci . The other groups had a smaller number of private alleles: four private alleles were found in Groups 2 and 4 , and only two private alleles were found in Group 3 ( Table 4 ) . Genetic differentiation based on the FST and Shannon’s mutual information index SHUA between groups estimated in the Structure program [50] exhibited similar but not identical patterns ( Table 5 ) . The highest differentiation values were found between pairs of groups 2–4 ( FST = 0 . 562 , SHUA = 0 . 203 ) , 3–4 ( FST = 0 . 470 , SHUA = 0 . 197 ) and 1–2 ( FST = 0 . 438 , SHUA = 0 . 270 ) . The lowest values of FST were obtained for groups 1 and 3 , although the lowest values of SHUA were indicated between groups 2 and 3 . Also , relatively low values of FST and SHUA were found between groups 1 and 4 ( 0 . 277 and 0 . 144 , respectively ) . Results of the Ewens-Watterson test indicated neutral evolution of all loci in all groups , but in Group 1 , in which neutrality was rejected in locus L43 as observed homozygosity values were lower than expected ( Table 6 ) . Because past demographic events can cause significant departures from neutrality , we tested Group 1 for the occurrence of possible bottleneck events . Results of the analysis performed with the program Bottleneck did not indicate any signs of recent population decline among R . nasutus specimens from Group 1 . One-tailed Wilcoxon sign-rank test for heterozygote excess indicated that each group was in mutation-drift equilibrium for all mutation models evaluated: IAM , TPM , and SMM ( P > 0 . 05 ) . EDENetwork analysis revealed strong connections between Groups 1 and 3 and 1 and 4 . Groups 2 and 3 were also connected . We did not observe any indication of substructuring within groups ( Clustering = 0 ) . Groups 1 and 3 were identified as possible source-sink populations based on the network analysis . These groups exhibited the highest connection values ( calculated through the degree parameter ) and betweenness , which indicate that they are exchanging migrants with other groups ( Fig 5A ) . Two of seven microsatellite loci ( L43 , List14-025 ) had complex repetition motifs that did not fit a SMM . Although List14-064 and List14-021 were reported with the same dinucleotide repeat motifs [43] , we observed nucleotide substitutions ( S2 Fig ) . Thus , we imposed the IAM mutation model as more appropriate for our data set for migration analysis . The estimates of the marginal likelihood ( lmL ) and the posterior model probability ( P ) for the 10 different models representing various scenarios of migration in Migrate-n revealed support for the 7th model ( Table 7 ) . Migration pattern of R . nasutus in the Caatinga biome inferred by Migrate-n analysis with “full migration” model ( Fig 5B ) revealed the existence of bidirectional gene flow between Groups 2 and 3 . Unidirectional gene flow was detected from Group 2 to 4 , from Group 1 to 3 , and from Group 4 to 3 . Group 1 seems to be a source population , because it only provides emigrants to other Groups , whereas Group 3 is a sink population as it is composed by immigrants from the other three populations . AMOVA between east ( Carnaúba dos Dantas , Jaguaruana , Sousa and Serra Talhada ) and west ( Altos , Campo Maior , Parnaíba and Piracuruca ) populations shows that the percentage of within-population variation is higher ( 53 . 18% , ФST = 0 . 46 , P < 0 . 0001 ) than among populations within groups ( 33 . 51% , ФSC = 0 . 39 , P < 0 . 0001 ) or between groups ( 13 . 31% , ФCT = 0 . 13 , P < 0 . 21 ) . The latter non-significant value of AMOVA statistic thus indicates gene flow between west and east groups . Linear regression analysis ( S3 Fig ) did not reveal a statistically significant correlation between cyt b-derived genetic distances and geographic distances ( R2 = 0 . 05 , P = 0 . 258 ) . However , a weak correlation was observed between microsatellite-based FST distances and geographic distances ( R2 = 0 . 29 , P = 0 . 003 ) , suggesting that distance contributes ~29% to the shaping of genetic diversity [63] .
The genetic structure revealed by microsatellite analyses ( Figs 4 and 5 ) indicate the existence of four highly differentiated groups with pairwise FST values ( >0 . 25; Table 5 ) well above those estimated for other triatomine populations at similar geographical scales , such as for R . prolixus in Venezuela [5] and T . infestans in Peru and Argentina [64 , 65] . Although this could indicate low levels of gene flow between populations or even the possibility of cryptic speciation ( as seen for other Rhodnius species [5] ) , results from the cyt b marker do not support the latter hypothesis . There is no evidence of present potential isolation by distance that could explain the population substructure observed herein . A weak correlation was found between genetic and geographical distances ( S2 Fig ) , assuming that the correlation between geographic and genetic distances indicate isolation by distance ( but see reference [66] for a discussion about the limitations of this analysis ) . Indeed , populations of Group 1 are geographically close and have low and non-significant pairwise FST values , but other populations such as Sousa and Carnaúba dos Dantas , which are also close , are highly structured ( FST = 0 . 51 , P < 0 . 01 ) . Therefore , we can hypothesize that restricted gene flow in the Caatinga region might result from human modifications , since the main sylvatic ecotope of this species , C . prunifera palm trees , has been reduced significantly due to continuous deforestation [67] . Biodiversity losses in tropical areas have been strongly determined by deforestation , which can interrupt environmental patches that uphold connectivity with other habitats [68] . Future research considering the possible routes of human migration among the studied communities , as well as a more complete analysis of the distribution of C . prunifera palm trees are necessary to adequately address this hypothesis . Reduced genetic diversity was detected in the three R . nasutus groups from eastern Caatinga ( groups 2 , 3 and 4 ) evidenced by the low number of alleles present ( < 5 ) in the majority of the loci analyzed , and low heterozygosity levels . Reduced diversity commonly occurs in endemic species which have low effective population sizes caused by habitat fragmentation [69] . In this case , stochastic processes such as genetic drift and inbreeding within isolated patches are expected to shape the genetic background of populations [70] . Indeed , Group 2 shows moderate levels of inbreeding ( FIS = 0 . 122 , P = 0 . 018 ) . Although two loci are monomorphic ( List14-025 and R31 ) , Group 1 from western Caatinga presents the highest heterozygosity levels . Moreover , this group has the highest Shannon’s allele information index SHA ( 0 . 955 ) , further indicating its high variability . Two loci display heterozygote excess ( L43 and List14-021 ) , which might result from isolate-breaking ( when two previously isolated populations are mixed [71] ) . It is important to note that Group 1 also has the highest number of private alleles . Perhaps the high number of private alleles observed in this group results from long-term isolation during glaciation periods . Evidence for departures from neutrality and bottleneck events were not detected in the R . nasutus populations analyzed . However , this result should be interpreted with caution , since the small sample size of some locations could prevent proper testing of this hypothesis [72] . Moreover , the reduction of effective population size , heterozygosity levels and allelic diversity may go undetected when ( i ) they occurred for a limited time interval or very recently [73] , ( ii ) are followed by a sudden population growth [74] , or ( iii ) when populations experienced low levels of subsequent immigration [75] . A scattered mosaic of habitat fragments , such as observed in the Caatinga region , might cause population reduction due to an impact in the connectivity of habitat networks , favoring dispersal through a series of successive stepping-stones events [76 , 77] . Network and Bayesian coalescent analyses indicated a western-eastern colonization of the biome . Group 1 is a source population , from which individuals emigrated to Groups 3 and 4 , while Group 3 is a sink population , also receiving emigrants from Groups 2 and 4 ( Fig 5A ) . Geographically close groups of eastern Caatinga have a complex pattern of gene exchange , with unidirectional migration of Group 2 individuals to Group 4 , and Group 4 individuals to Group 3 , as well as bidirectional gene flow between Groups 2 and 3 ( Fig 5B ) . Unidirectional gene flow may be explained by a decrease in capacity at the source habitat and availability of expansion potential at the target area [78] . The niche capacity of R . nasutus may be limited by several ecological factors , such as the sparse distribution of preferred food sources ( fauna associated with the palms ) and human activity . The network shows that most haplotypes differ by a single substitution , indicating reduced diversity . These haplotypes seem to be under neutral evolution , since 13 out of 16 variable sites resulted in synonymous substitutions . Two haplotypes are shared by five geographically distant locations ( Serra Talhada , Parnaíba , Jaguaruana , Campo Maior and Altos ) , which could indicate high gene flow among populations or reflect retention of ancestral polymorphism [79] . The remaining 14 haplotypes are exclusive ( Fig 2 ) , and thus support restricted migration . Pairwise FST comparisons corroborate the latter hypothesis , as 24 out of 28 values are high ( > 0 . 6 ) and significant ( Table 3 ) . Populations can become structured quickly during a range expansion event as a consequence of the colonization of newly available patchy areas by pioneering groups with reduced population sizes [76] . In the case of a patchy environment , there are limitations to dispersal and therefore the colonization of new habitats will increase genetic structure among populations as only a subset of the original gene pool will be transferred [72] . Evidence suggests that populations of R . nasutus expanded from west to east throughout the Caatinga in the last 12–10 ka . Molecular diversity indices showed high haplotype diversity ( Hd ) and low nucleotide diversity ( π ) , which indicates that despite the high number of different haplotypes per population , they only differ by 1–5 nucleotides . This observation is consistent with expansion of source populations with small effective size [79] . Mismatch distribution analysis showed a good fit to the expected under the model of population growth ( Fig 3A ) . Moreover , the demographic model that better represents the population dynamics of R . nasutus is the exponential growth ( BF = 25 . 11 ) , evidenced by a 5-fold increase in the species population size in the Holocene ( Fig 3B ) . The demographic history of R . nasutus is estimated to have begun 66 thousand years ago thus covering Pleistocene–Holocene geological epochs ( Fig 3B ) . Based on the Bayesian Skyline Plot results , the population size of R . nasutus was stable for approximately 50 , 000 years . This demographic stability may reflect the presence of a putative refuge area for Seasonally Dry Tropical Forests in the Caatinga , spanning most of the species current distribution area during the Last Glacial Maximum ( 24–12 ka ) [16] . R . nasutus went through a population expansion 12–10 ka , during the Holocene . This expansion is consistent with global climate changes ( Fig 3B ) , as it coincides with the end of the largest dry season in South America , during the glaciation Wisconsin-Würm ( 18–12 ka ) . This period is characterized by the expansion of semi-humid vegetation in South America and diversification of Neotropical biota in seasonally dry areas [80 , 81] . Moreover , the Caatinga biome experienced an increase in moisture levels during the late Pleistocene ( 12–6 ka ) , which possibly contributed to the expansion of Mauritia palm tree species in the region [82] and likely also Copernicia palms . The genetic structure of R . nasutus populations inferred based on the two markers differed considerably ( Figs 2 and 4 ) . Cyt b analysis indicated more complex substructure than the microsatellite data ( Table 3 ) . Although Group 3 ( Serra Talhada ) and 4 ( Sousa ) remained separate by both microsatellites and cyt b data , microsatellite-defined Group 1 comprises four populations ( Altos , Campo Maior , Piracuruca and Parnaíba ) , and Group 2 includes two populations ( Jaguaruana and Carnaúba dos Dantas ) , whereas cyt b data grouped together Altos and Campo Maior with the geographically distant population of Jaguaruana . Assuming that sampled populations represent panmitic units , we might speculate that in the past , Altos , Campo Maior and Jaguaruana were connected , while Piracuruca and Parnaíba were separated . In a contemporary event , the Jaguaruana population became isolated from the western group , which clustered with the other two western populations , Piracuruca and Parnaíba . It is worth mentioning , however , that FST results based on cyt b data should be interpreted with caution , due to the low divergence found between R . nasutus sequences . In a contemporary event , the Jaguaruana population became isolated from the western group , which clustered with the other two western populations , Piracuruca and Parnaíba . Group 1 heterozygote excess and pairwise cyt b and microsatellite FST values among populations ( 0 . 20–0 . 77 and 0 . 07–0 . 24 , respectively ) , thus corroborate the hypothesis of an isolate-breaking effect . Incongruent results may also be explained by different evolutionary mechanisms ( inheritance modes , population size , rates of evolution , response to selective pressure ) and intrinsic biological peculiarities of triatomines ( e . g . males fly more than females ) . Poor accordance between population genetic structures inferred from mitochondrial and microsatellite markers have also been reported for various animal groups , including mammals [83] , amphibians [84] and insects [85] . R . nasutus was able to recently disperse within the Caatinga in response to favorable climatic events and environmental conditions . The species is presently facing an important ecological impact caused by the continuous deforestation of C . prunifera palm trees . Perhaps this ecological disturbance is contributing to the observation that R . nasutus now colonizes other species of palms ( A . speciosa , M . flexuosa , S . oleracea , A . intumescens [12 , 13] ) and occasionally , even trees ( L . rigida [14] ) . These observations show that R . nasutus populations have the ability to adapt to new microhabitats . In addition , it also invades human-modified areas and sporadically colonizes dwellings [10 , 11 , 86] . We observed that C . prunifera palm trees in deforested areas seem to be more prone to become infested with R . nasutus than those in forested areas ( 87% and 58% , respectively ) , which raises the risk of T . cruzi transmission to humans . Demographic history data indicates the capacity of R . nasutus to increase its population size when environmental conditions are favorable . Therefore , closer attention should be paid to the distribution of R . nasutus , as it can maintain T . cruzi transmission within the Caatinga . Given the limitations inherent to insecticide spraying such as continuous re-infestation of insecticide-treated households by abundant native vectors , molecular studies focusing on the distribution patterns and demographic trends of sylvatic triatomines should be conducted in order to further improve the effectiveness of Chagas disease control . | Chagas disease is endemic to Latin America and the Caribbean and it is estimated that 6–7 million people are infected with the etiological agent Trypanosoma cruzi . Although new community-based ecosystem management ( ecohealth ) initiatives have been implemented , vector control based on insecticide-spraying of households remains one of the most effective strategies to diminish parasite transmission to humans . However , this strategy is not sustainable where native triatomine species are capable of colonizing peridomestic structures and invading human dwellings . The application of molecular markers with the potential of recovering both historical and contemporary information on vector population structure and diversity can improve the understanding of vector dissemination and thus contribute to the development of better disease control strategies . In this study we analyzed Rhodnius nasutus populations endemic to the Brazilian Caatinga biome using two sets of markers: a fragment of the mitochondrial gene cytochrome b and eight nuclear microsatellite loci . The information generated and described herein is important as it may contribute to the advancement of our understanding of Chagas disease vector ecology and phylogeography . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"biogeography",
"taxonomy",
"invertebrates",
"ecology",
"and",
"environmental",
"sciences",
"medicine",
"and",
"health",
"sciences",
"population",
"genetics",
"animals",
"genetic",
"mapping",
"animal",
"models",
"phylogenetics",
"data",
"management",
"phylogenetic",
"anal... | 2018 | Phylogeography and demographic history of the Chagas disease vector Rhodnius nasutus (Hemiptera: Reduviidae) in the Brazilian Caatinga biome |
On its own , a single cell cannot exert more than a microscopic influence on its immediate surroundings . However , via strength in numbers and the expression of cooperative phenotypes , such cells can enormously impact their environments . Simple cooperative phenotypes appear to abound in the microbial world , but explaining their evolution is challenging because they are often subject to exploitation by rapidly growing , non-cooperative cell lines . Population spatial structure may be critical for this problem because it influences the extent of interaction between cooperative and non-cooperative individuals . It is difficult for cooperative cells to succeed in competition if they become mixed with non-cooperative cells , which can exploit the public good without themselves paying a cost . However , if cooperative cells are segregated in space and preferentially interact with each other , they may prevail . Here we use a multi-agent computational model to study the origin of spatial structure within growing cell groups . Our simulations reveal that the spatial distribution of genetic lineages within these groups is linked to a small number of physical and biological parameters , including cell growth rate , nutrient availability , and nutrient diffusivity . Realistic changes in these parameters qualitatively alter the emergent structure of cell groups , and thereby determine whether cells with cooperative phenotypes can locally and globally outcompete exploitative cells . We argue that cooperative and exploitative cell lineages will spontaneously segregate in space under a wide range of conditions and , therefore , that cellular cooperation may evolve more readily than naively expected .
Many cell phenotypes alter the growth and division of nearby cells by changing local resource availability [1]–[4] . Some of these phenotypes promote the survival and reproduction of others , and thus qualify as a simple form of cooperation . A cell may be considered cooperative , for example , if it secretes enzymes that free nutrients which neighboring cells can use . The efficiency with which a cell group processes environmental resources or exploits a host often depends on such publicly beneficial cell phenotypes . For instance , many microbial infections and cancerous tumors derive their pathogenicity in part from the cooperative secretion of digestive enzymes by their constituent cells [5]–[8] . How cooperative cell phenotypes evolve therefore presents an important question , one that is particularly challenging because any genetic variants that exploit others' cooperation – without themselves paying a cost – can potentially invade and increase in frequency . In light of this problem , social evolution theory has been developed to understand the evolutionary trajectories of cooperative traits [9] , but this framework has only recently been applied to unicellular systems [4] , [10]–[12] . The critical prediction is that preferential interaction among genetically related individuals increases the propensity for cooperative phenotypes to evolve . Variation among individual cells is a common feature of many cell groups: microbial biofilms are often composed of multiple strains or species [13] , [14] , and cancerous tumors can consist of many different genetic lineages [15] , [16] . The majority of work on cooperative cell phenotypes assumes relatively well mixed interactions among different genetic variants in standing or shaken liquid culture [17]–[21] . This kind of environment does not reflect the natural condition of most cell groups , in which cells are typically constrained in space and influence each other in a distance-dependent manner . These spatial relationships may be paramount to understanding the evolution of cellular cooperation [22] . When different cell lineages are segregated in space , those expressing cooperative phenotypes are more likely to benefit others of their own kind [23]–[25] . When different cell lineages are mixed together , on the other hand , cells that exploit the resources of others can thrive [17]–[20] . Local populations of bacterial and cancer cells are often established by groups of progenitors that proliferate into larger clusters . Experiments with bacterial colonies on agar have revealed that expanding cell groups can segregate into sectors that are each dominated by a single genetic lineage [26] , [27] . This observation has been used predominantly to motivate new population genetic models [28]–[30] . When only cells on the periphery of an expanding group can access nutrients and reproduce , the group's effective population size is reduced . As a result , neutral or even mildly deleterious alleles can spread by genetic drift along the advancing front . Because they are constrained in space , genetic lineages that manage to proliferate along the population's leading edge become physically separated into zones composed of clonal or closely related individuals . By promoting interaction between individuals of the same genotype , the spontaneous segregation of different genetic lineages in space may also influence social evolution within cell groups [23] , [24] . In the present paper , we use a generalized mechanistic model to define the physical and biological factors that govern cell group spatial structure , and we explore the potential connection between genetic drift along the fronts of expanding cell groups and the evolution of social phenotypes .
We began with simulations in which the environment surrounding cell groups was altered by increasing or decreasing growth substrate concentration . These in silico experiments were initiated with equal numbers of randomly distributed red and blue cells , which did not differ in any way other than their color . The two neutral color markers were used to judge whether cell lineages remain randomly mixed or become spatially segregated as cell groups expand . Environmental substrate availability was decreased from saturating to sparse across multiple simulations , and we observed three different regimes in cell group structure: Our next goal was to describe why environmental substrate concentration affects lineage assortment in expanding cell groups . Under limited growth substrate availability , the majority of cell growth and division occurs along a group's advancing front in an active layer whose depth depends on substrate penetration ( Figure 3 ) . Previous work has hinted that active layer depth is a critical factor influencing cell group surface structure [39] , [40] , and we therefore hypothesized that it is not substrate concentration in particular , but more generally the depth of a cell group's active layer that controls cell lineage segregation . Because segregation increased as growth substrate supply decreased in our preliminary simulations , we predicted that thinner active layers would lead to stronger lineage segregation in expanding cell groups . Active layer depth is not solely a function of bulk growth substrate concentration . For example , higher substrate diffusivity increases active layer depth by allowing substrate to enter further into the cell group before being depleted . Faster cell growth rates , on the other hand , decrease active layer depth by raising the rate of substrate consumption at the cell group's outer surface . If we are correct that active layer depth is the underlying determinant of lineage segregation , all of the physical and biological factors that control active layer depth should also influence lineage segregation in cell groups . Using an analytical technique from chemical engineering ( Methods ) , we combined the factors that influence active layer depth into a dimensionless number , δ , which has the following form for our system: ( 1 ) Here , Gbulk is the bulk liquid concentration of growth substrate , DG is the growth substrate diffusion coefficient , Y is the yield with which cells convert substrate to biomass , μmax is the maximum specific cell growth rate , ρ is the cell biomass density , and h is the height of the diffusion boundary layer ( Figure 3 ) . The smaller the value of δ , the thinner the cell group's active layer . We performed three new sets of simulations to test the hypothesis that active layer depth controls cell lineage segregation . Within each set , we varied active layer depth ( δ ) by altering only one parameter from Equation 1: maximum cell growth rate ( μmax ) , bulk growth substrate concentration ( Gbulk ) , or growth substrate diffusivity ( DG ) . At the end of each simulation , we calculated the segregation index . Our hypothesis makes two key predictions: 1 ) cell lineage segregation should be inversely related to δ , a proxy for active layer depth . 2 ) The relationship between cell lineage segregation and δ should be independent of which parameter from Equation 1 is altered . The results are shown in Figure 4 and support both predictions . Lineage segregation within cell groups declines with increasing δ , regardless of how δ is altered . Using the dimensionless number δ renders our results independent of the exact values of Gbulk , DG , Y , μmax , ρ , and h used to run simulations . It is the relative magnitudes of these parameters in combination that ultimately matter . How does active layer depth influence cell lineage segregation ? When growth substrate penetrates through most of a cell group before being depleted , all cells grow and divide , pushing each other into a homogeneous mixture . As active layer depth decreases below the total thickness of a cell group , however , cells that happen to fall below a critical distance from the group's front can no longer contribute to population expansion . Decreasing active layer depth thus reduces the cell group's effective population size , rendering it more susceptible to genetic drift along its advancing front . Because the cells are constrained in space , reductions in genetic diversity along the group's leading edge lead to localized clusters of individuals that all descend from a common progenitor [30] . This phenomenon – often referred to as sectoring or gene surfing [28]–[30] – has been observed in agar colonies of Paenibacillus dendritiformis [26] , Escherichia coli and Saccharomyces cerevisiae [27] . Reducing active layer depth even further yields an additional qualitative shift in cell group structure: the expanding population becomes sensitive to small irregularities along its leading edge . Cells in the peaks of surface irregularities retain access to substrate and grow into tower projections , while cells in the troughs of surface irregularities lose access to substrate and cease growing . This process is related to viscous fingering at the interface of two fluids [39] , [41] , and it is known to generate rough surface structure along the leading edges of growing biofilms , bacterial colonies on agar [34] , [35] , [40] , and moving fronts in general [36] . From a biological perspective , our analysis predicts that such surface roughness is accompanied by abrupt genetic lineage segregation along the front of an expanding population . The spatial assortment of cell lineages is potentially critical for traits that affect the reproduction of other individuals in the population . It is increasingly recognized that cells express many such social phenotypes [4] , [12] , which are often involved in nutrient acquisition and pathogenesis [42]–[45] . A common example is the secretion of extracellular enzymes or nutrient-chelating molecules . Cells that synthesize these substances must forgo a fraction of their reproductive capacity [17]–[19] , but if enough cells participate , all can gain a net benefit ( to the detriment of their host , in the case of pathogens ) . In many cases the evolution of simple cooperative phenotypes depends on three factors: 1 ) c , the cost incurred by cooperative individuals 2 ) b , the benefit gained by the receivers of cooperative behavior , and 3 ) r , the correlation between genotypes of givers and receivers of cooperation . Cooperation is predicted to evolve when rb>c , a condition known as Hamilton's Rule [9] . The cost and benefit factors are measured in terms of reproductive fitness . When cooperation is genetically determined , relatedness may be thought of as the degree to which the benefits of cooperation are preferentially distributed to other cooperative individuals . The segregation index depicted in Figures 2 and 4 is equivalent to a form of the relatedness coefficient in Hamilton's Rule: both measure the degree of biased interaction among relatives ( here , physical proximity amounts to biased interaction ) . As such , our segregation index forms a bridge between social evolution theory and the emergence of lineage segregation in cell groups , allowing us to extend our prediction from the previous section . Because thin active layer conditions generate lineage segregation , we predict that decreasing active layer depth will promote interaction among clonemates ( increasing r in Hamilton's Rule ) and favor the evolution of cooperation [9] , [12] , [23] . Positive spatial assortment of related cells does not guarantee that cooperation will be favored , however , as the same segregation that allows cooperators to preferentially interact also increases the strength of competition between them [24] . We tested our prediction by implementing a cooperative phenotype in our model framework and competing cooperative cells against exploitative cells that devote all resources to growth . Cooperative individuals secrete a diffusible compound that benefits all other cells in the local area ( we will refer to the compound as an extracellular enzyme ) . Local availability of the secreted enzyme increases cell growth rate by a fold factor B , but only after the enzyme's concentration passes a threshold value , τ . Cooperative cells constitutively secrete the enzyme and incur a fold decrease in growth rate of C x RE , where C is a cost scaling factor and RE is the enzyme production rate . In our main analysis , B = 3 , C = 0 . 3 , and RE ranges from 0 to 2 . We derived these values from experimental data on elastase , a secreted enzyme and virulence factor of the bacterial pathogen Pseudomonas aeruginosa [19] , [46] . We asked whether a cooperative cell line , which pays a cost to produce a diffusible , publicly beneficial compound , could outcompete an exploitative cell line that invests all of its resources into growth . Each competition simulation began with a randomly distributed 1∶1 mixed monolayer of the two cell types , and cell groups were grown to a maximum height of 100 µm . We then calculated the evolutionary fitness of the cooperative cell line , relative to that of the exploitative cell line ( Methods ) . This competition pairing was repeated over a range of extracellular enzyme production rates on the part of cooperative cells . The higher the enzyme production rate , the more rapidly cells accrue its benefit , but the larger the cost suffered by cooperative cells . Finally , all competition pairings were repeated across three active layer depth conditions ( δ = 10 , 2 , 1 ) , representing the three cell group structure regimes described in Figure 1 . Figure 5 summarizes the results of our competition simulations . When active layers are thick ( δ = 10 ) , leading to well mixed cell lineages , the extracellular enzyme is homogenously distributed through cell groups . The non-cooperative cell line is therefore able to consistently exploit and outcompete the cooperative cell line ( Figure 5A ) . This result is consistent with numerous observations that exploitative mutants outcompete enzyme-secreting bacteria when they are inoculated together in liquid culture , in which cell lineages largely remain mixed [17]–[20] . When active layer depth is decreased ( δ = 2 ) , there is a narrow range of extracellular enzyme production rates at which cooperative cells outcompete exploitative cells ( Figure 5B ) . The critical difference is that cooperative cells and exploitative cells no longer remain well mixed; rather , they segregate into clonal regions . As a result , the benefit of extracellular enzyme released by cooperative cells accrues asymmetrically to other cooperative cells . The range of enzyme production rates at which cooperative cells prevail is narrow , however , because the benefits of lineage segregation ( increasing r in Hamilton's Rule ) can be outweighed by the cost of higher extracellular enzyme production ( increasing c in Hamilton's Rule ) . Further decreasing active layer depth ( δ = 1 ) leads to the growth of spatially isolated , clonal cell towers . Under these conditions , the benefits of a cooperative secreted enzyme are distributed even more asymmetrically to other cooperative cells . Consistent with our predictions , this allows cooperative cells to outcompete exploitative cells over a larger range of enzyme production rates ( Figure 5C ) . We also noted the sizable variation between simulation runs when δ = 1 , particularly if extracellular enzyme production rates were low ( Figure 5C , enzyme production rate = 0 , 0 . 25 , 0 . 5 ) . This variation reflects a founder effect; it manifests most strongly when there is no or little difference between the competitive abilities of cooperative and exploitative cell lines , rendering the outcome of each simulation subject to chance events that determine which cells seed the few tower structures that emerge from an expanding cell group . Our results show that thin active layer conditions allow cells expressing cooperative phenotypes to outcompete exploitative cells within a single cell group . To better account for the long-term evolution of a metapopulation comprising many cell groups , we performed an invasion analysis to determine whether a novel cooperative mutant can spread through a metapopulation otherwise containing only exploitative cells ( Supporting Information , Text S1 ) . We also examined the reciprocal case to determine if a rare exploitative mutant can invade a metapopulation otherwise containing only cooperative cells [32] , [33] . We found that cooperation can invade under a large swath of parameter space , but only under thin active layer conditions that promote lineage segregation can cooperative cells eliminate exploitative cell types on a metapopulation scale ( Supporting Information , Figure S2 ) . The results of both our local competition and invasion analyses are robust to the cost/benefit ratio of cooperation , with one partial exception when cells invest very heavily into an expensive cooperative phenotype ( Supporting Information , Figure S3 ) . Our study indicates that an order of magnitude change in nutrient availability , nutrient diffusivity , cell metabolic efficiency , cell growth rate , or biomass density can shift cell groups from a regime of lineage mixing to a regime of pronounced lineage segregation . The number δ defined in Equation 1 relates these parameters to the depth of a cell group's active layer , which governs how cell lineages become spatially assorted over time . Thick active layers promote lineage mixing , while decreasing active layer depth generates increasingly strong lineage segregation . Cell lineage segregation , in turn , favors the evolution of cooperative phenotypes . Previous work performed with bacteria in liquid planktonic culture has concluded that cooperative cell phenotypes cannot be selectively favored within a single population also containing exploitative cells [17] , [19] , [20] . Our study shows that this conclusion will not always hold because cooperative cells can spontaneously segregate from exploitative cells when they are constrained in space . Our results also imply that , given realistic parameters for a cooperative cell phenotype , the benefits of preferential interaction between cooperators can outweigh the costs of increased competition between related cells that are clustered together in space [24] . Like all models , ours uses simplifying assumptions . We deliberately omit some physical processes , such as shear stress , that may be applied to cell groups in the real world [47] . Our simulations also do not consider active cell motility , which in reality could influence cell group structure and evolution . We have additionally assumed that cell phenotypes of interest , such as extracellular enzyme secretion , are expressed constitutively or not at all . In nature , the expression of many social phenotypes is adjusted in response to environmental cues [48]–[50] . Though these simplifications should be assessed theoretically and empirically , they were critical in allowing us to identify basic physical and biological parameters that control cell group structure and evolution . In summary , our model suggests that clusters of genetically related cells can emerge quite easily in spatially constrained cell groups , even when cells possess no mechanism for actively gathering with clonemates . Lineage segregation allows cooperative cells to outcompete exploitative cells , and accordingly we predict that localized cooperation will evolve more readily in cell groups than suggested by models and experiments that only consider liquid environments .
We simulate cell groups using an individual-based model described in detail previously [31] . Simulation parameters are listed in Table S1 ( Supporting Information ) . Cell growth is a function of the local microenvironment , namely the concentrations of solutes such as growth substrate ( G ) and extracellular enzyme ( E ) ( Supporting Information , Table S2 ) . The uptake of growth substrate by each cell is considered when calculating the spatial gradients of substrate concentration . We achieve this by solving a reaction-diffusion equation , where r is a growth rate expression: ( 2 ) Following the common assumption that reaction-diffusion is much faster than cell growth and division [31] , our simulations proceed according to the following steps: The individual-based simulation framework was written in the Java programming language , and its related numerical methods are detailed elsewhere [31] . Briefly , they include the Euler method to grow cells at each iteration , a hard-sphere collision detection method to identify pushing events between neighboring agents , and the FAS multigrid to solve reaction-diffusion equations to steady state [51] . The 3D images in Figure S1 where rendered using POV-Ray . All other figures were prepared using Matlab ( the Mathworks , Inc . ) . The computations in this paper were run on the Odyssey cluster supported by the Harvard University FAS Research Computing Group . To obtain the segregation index for a cell group at a single point in time , we first identify every actively growing cell . These M cells are indexed by Ai: A1 , A2 , … , AM . To measure segregation with respect to a single focal cell Ai , we identify all other individuals within a distance of 10 cell lengths . The N cells in this neighborhood are indexed by aj: a1 , a2 , … , aN . We define a genetic identity function , g ( aj ) : ( 3 ) and a metabolic activity function , m ( aj ) : ( 4 ) where [G] is the local concentration of growth substrate , and KG is the half-saturation constant for cell growth rate . Segregation with respect to a focal cell , s ( Ai ) , is calculated as the mean product of the g and m functions for every cell in its neighborhood: ( 5 ) Finally , we define the segregation index for the entire cell group as the mean value of s ( Ai ) across all metabolically active cells: ( 6 ) Our segregation index measures the degree to which co-localized , metabolically active cells are clonally related to each other . The index is equal to a form of the relatedness coefficient from social evolution theory under the following assumptions: 1 ) A cell expressing the cooperative phenotype equally benefits all other individuals within a 10 cell-length radius; 2 ) Each cell within range of receiving cooperative benefits makes a contribution to mean relatedness proportional to its growth rate; 3 ) Cell groups are seeded randomly from a large population pool . The dimensionless number , δ , is a proxy for the depth to which growth substrate penetrates into a cell group before being depleted by cell metabolic activity . δ is derived by non-dimensionalizing Equation 2 . We normalize growth substrate concentration by its bulk liquid concentration , , and local biomass by cell biomass density , x = X/ρ . We then normalize the space coordinates by the height of the boundary layer , h . The steady state , dimensionless version of Equation 2 becomes: ( 7 ) Note that the factor multiplying the Laplacian of , , is the square of δ as defined in the main text . δ is also the inverse of the Thiele modulus [52] , a number commonly used in chemical engineering to quantify the activity of solid catalysts . We calculate the competitive fitness of each cell line as the mean number of rounds of cell division per unit time that each achieves over the course of a simulation: ( 8 ) where NS , t is the number of cells of strain S present within the cell group at time t . The relative fitness of a strain S1 in local competition with another strain S2 is defined as: . | Cooperation is a fundamental and widespread phenomenon in nature , yet explaining the evolution of cooperation is difficult . Natural selection typically favors individuals that maximize their own reproduction , so how is it that many diverse organisms , from bacteria to humans , have evolved to help others at a cost to themselves ? Research has shown that cooperation can most readily evolve when cooperative individuals preferentially help each other , but this leaves open another critical question: How do cooperators achieve selective interaction with one another ? We focus on this question in the context of unicellular organisms , such as bacteria , which exhibit simple forms of cooperation that play roles in nutrient acquisition and pathogenesis . We use a realistic simulation framework to model large cell groups , and observe that cell lines can spontaneously segregate from each other in space as the group expands . Finally , we demonstrate that lineage segregation allows cooperative cell types to preferentially benefit each other , thereby favoring the evolution of cooperation . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"evolutionary",
"biology",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/pattern",
"formation"
] | 2010 | Emergence of Spatial Structure in Cell Groups and the Evolution of Cooperation |
Since 1999 a lineage of the pathogen Cryptococcus gattii has been infecting humans and other animals in Canada and the Pacific Northwest of the USA . It is now the largest outbreak of a life-threatening fungal infection in a healthy population in recorded history . The high virulence of outbreak strains is closely linked to the ability of the pathogen to undergo rapid mitochondrial tubularisation and proliferation following engulfment by host phagocytes . Most outbreaks spread by geographic expansion across suitable niches , but it is known that genetic re-assortment and hybridisation can also lead to rapid range and host expansion . In the context of C . gattii , however , the likelihood of virulence traits associated with the outbreak lineages spreading to other lineages via genetic exchange is currently unknown . Here we address this question by conducting outgroup crosses between distantly related C . gattii lineages ( VGII and VGIII ) and ingroup crosses between isolates from the same molecular type ( VGII ) . Systematic phenotypic characterisation shows that virulence traits are transmitted to outgroups infrequently , but readily inherited during ingroup crosses . In addition , we observed higher levels of biparental ( as opposed to uniparental ) mitochondrial inheritance during VGII ingroup sexual mating in this species and provide evidence for mitochondrial recombination following mating . Taken together , our data suggest that hypervirulence can spread among the C . gattii lineages VGII and VGIII , potentially creating novel hypervirulent genotypes , and that current models of uniparental mitochondrial inheritance in the Cryptococcus genus may not be universal .
Cryptococcus neoformans and C . gattii are the causative agents of cryptococcosis in humans . C . neoformans typically infects HIV-infected individuals and other patients with immunodeficiencies , but has also been found in apparently immunocompetent individuals in the Far East [1] , [2] . C . gattii is a primary pathogen that causes disease in otherwise healthy people [3] , [4] , but has also been found in HIV patients in Malawi , Africa and California , USA [5] , [6] . C . gattii accounts for less than 1% of all cryptococcosis cases , and until the late 1990s occurred mostly in subtropical regions of the world . However , in 1999 , an outbreak of C . gattii was reported on Vancouver Island in domestic pets and people [7]–[9] . This outbreak spread to mainland Canada and then into the northwestern states of the United States [10]–[13] and currently numbers more than 400 cases [14]–[17] . C . gattii is divided into distinct clades ( VGI-VGIV ) [14] , with the outbreak originating on Vancouver Island , and a more recent outbreak in Oregon [18] , , being caused by three clonal groups within VGII ( VGIIa , VGIIb and VGIIc ) [20] . These hypervirulent outbreaks are characterized by an unusual ability of the pathogen to parasitise host phagocytic cells: upon engulfment by macrophages , outbreak strains initiate mitochondrial tubularisation and rapid intracellular proliferation of the fungus [21] . Cryptococcosis is not spread from infected animals or humans to susceptible hosts but rather infections are acquired from the environment . Hence , cryptococcal species likely experience strong selective pressure from factors encountered within environmental niches . Genetic recombination by meiotic sexual reproduction in eukaryotic pathogens is a widely-occurring mechanism that generates genetic diversity ( and hence novel phenotypic diversity ) but carries the risk of destroying beneficial gene combinations [22] . The genetic distance across which genetic recombination occurs yields very different outcomes . Outcrossing and hybridization can result in dramatic changes to genotype and resulting virulence phenotypes . For example , Grigg et al . [23] have demonstrated that outcrossing sexual recombination can be a major force in shaping eukaryotic pathogens , since recombinant Toxoplasma progeny from crosses between two distinct ancestral lines type II and type III are significantly more virulent than either parent . A similar hypothesis has been proposed for the origin of C . gattii outbreak strains [24] . However , outcrossing can also come at the cost of breaking up highly-fit coadapted gene-complexes , such as those that enable host adaptation [25] , [26] , and can result in lethal levels of genetic load resulting in widespread inviability . Therefore , estimating how likely it is for hypervirulence traits to move between C . gattii lineages by recombination is critical both for predicting the likelihood of novel hypervirulent genotypes occurring and , more broadly , in understanding the origins of infectious outbreaks . In addition , given that the expression of mitochondrially-encoded genes correlates with virulence in C . gattii , but mitochondrial genes do not contribute to virulence in C . neoformans [21] , [27] , important questions remain about the relative role of mitochondrially-encoded , versus nuclear-encoded , genes in controlling virulence in this pathogen . Here we address both of these questions via a series of genetic crosses , followed by comprehensive phenotypic analyses . Our findings demonstrate that hypervirulence in C . gattii is a complex , multigenic trait . Surprisingly , however , this trait can be transmitted relatively easily to other lineages and is not strictly limited to one mitochondrial genotype . Finally , we show that , in contrast to existing paradigms , mitochondrial inheritance in C . gattii is not strictly uniparental and thus current models of genetic exchange in this pathogenic clade should be revisited .
This study addresses two questions: Recent work has demonstrated that the ability to change mitochondrial morphology is closely linked to intracellular proliferation and thus hypervirulence in C . gattii [21] . In the related pathogen C . neoformans , as in most eukaryotes , mitochondria are inherited from only one parent ( in this case the MATa parent ) following mating [28] , [29] . To exploit this uniparental inheritance and to test the likelihood of virulence traits spreading within the C . gattii population , we conducted a series of crosses in which progeny would inherit mitochondria either from a hypervirulent parent , or from a non-outbreak strain exhibiting wild type virulence . If phenotypes associated with hypervirulence ( mitochondrial tubularisation in response to phagocytosis and rapid intracellular proliferation ) were solely determined by mitochondrial genotypes then all progeny from each cross would have the same phenotype as the MATa parent ( Figure 1A ) . The ability to proliferate within macrophages is a proven predictor of virulence in C . neoformans and C . gattii [5] , [21] . To assess virulence in a comprehensive progeny set in this study , we utilized intracellular proliferation as a proxy-measure of virulence and investigated its relationship to mitochondrial tubularisation . Our experimental approach included both outgroup crosses , between strains from two different molecular groups ( VGII and VGIII ) , and ingroup crosses , between strains from the same molecular group ( VGII ) ( Figure 1B ) . Despite the fact that VGIII strains are more fertile than other C . gattii strains [24] , [30] , [31] , experimental mating of C . gattii strains is extremely difficult in the laboratory setting . However , within the VGIII lineage , the VGIII pair B4546 ( MATa ) and NIH312 ( MATα ) had previously been identified as mating test strains in an extensive screening study [32] . These strains also exhibit low intracellular proliferation rates and hence were chosen for the outgroup crosses in this study . Disappointingly , after various attempts , we and others were unable to mate MATa-VGII with MATα–VGII strains that exhibit explicitly distinct intracellular proliferation values . We were , however , able to conduct mating between VGII strains with more similar intracellular proliferation rates allowing for dissection of individual spores . Overall , there was a low rate of spore germination in both the VGII a x VGIII α and VGII α x VGIII a mating pairs . This is consistent with population genetic evidence [24] , [33] that these molecular types are genetically isolated with respect to nuclear DNA exchange and hence consistent with assignment as distinct species [34] . For each mating ( B4546 x R265 , CBS10090 x NIH312 and JF101 x AIg289 ) , at least 50 individual spores ( 50 , 50 , 63 ) were dissected , and in all cases none germinated ( 0/163 ) . After extensive attempts we were able to obtain six viable microdissected spores ( 6/140 ) from an outgroup cross ( strains YL4 x 97/433 ) . Crosses across species boundaries in Escherichia , Salmonella and Saccharomyces species are known to suffer from extensive DNA mismatches and cause serious problems during meiosis attributable to the mismatch repair system aborting homologous recombination [35]–[37] . To circumvent this substantial barrier and in an attempt to generate a more comprehensive working progeny set , a region of highly dense spores and hyphae was selected , plated , and colonies that arose were isolated and characterized . This type of analysis is therefore subject to possible isolation of parental yeast cells , blastospores ( yeast cells derived from hyphae post-fusion but prior to nuclear fusion and meiosis ) , diploid fusion products , and true haploid meiotic progeny . For these reasons , and to better understand the dynamics of mating between these two distinct molecular types , the resulting progeny sets were subjected to molecular ( MultiLocus Sequence Typing; MLST/Fluorescence-activated cell sorting; FACS ) and phenotypic ( self-filamentation ) analyses . As an additional approach , VGIII strains carrying the crg1::NEO mutation were crossed with a VGII strain carrying the bwc2::NAT mutation ( JF101xAIg289 and JF109xAIg254 ) ( Figure 1B ) . The parental strain AIg289 also carried a mutation within the FUR1 gene that confers resistant to 5-fluorouracil . Mutation of both Crg1 and Bwc2 enhance mating under conditions that normally repress it . Basidiospores were unviable ( 0/63 germinated ) . However , putative fusion or post-meiotic strains were isolated by plating onto YPD medium containing both nourseothricin and G418 . Strains were examined for phenotypes and PCR-RFLP markers . In the mating pair between VGII α ( R265 ) x VGIII a ( B4546 ) , we isolated a total of 18 progeny as described above . Amplification of the ATP6 gene ( encoded by the mitochondrial genome ) revealed that 100% of the progeny inherited the a mitochondrial genome ( Figures 2A&B and Supplementary Information ) , consistent with previous studies showing uniparental mitochondrial inheritance during a-α mating [28] , [38] , [39] . Based on FACS analysis , we then determined that 18/18 isolates showed signs of diploidy ( Figure 2 C&D ) and , in line with this , MLST analysis showed signs of heterozygosity at 7/8 markers ( Figure 2C ) . In all of the isolates , the remaining marker ( PLB1 ) specifically amplified the VGII allele , most likely due to primer bias or loss of heterozygosity caused by mitotic gene conversion or partial chromosomal loss . Interestingly , while all 18 strains retained copies of both SXI1/SXI2 , only 17% ( 3/18 ) were self-fertile ( Figure 2D ) . Six progeny that were restored towards haploidy following extensive passage on YPD medium , and these showed recombinant genotypes in MLST analysis with alleles contributed by both parental strains , and ploidy was assessed by FACS analysis ( Figure 2E ) . None of the progeny showed signs of self-filamentation , although none were derived from one of the three self-filamentous parents ( Figure 2E ) . It is problematic to distinguish aneuploid progeny by FACS analysis alone . MLST analysis showed signs of heterozygosity for the markers MPD , GPD1 and LAC1 for all restored “haploid” progeny . We therefore sequenced the genomes of four of the diploid progeny as well as the respective “haploid restored” progeny and examined read mapping coverage and variant ratios to determine the ploidy in these strains . This analysis revealed that all progeny were broadly diploid/haploid , but most also carried aneuploid regions within at least one of their chromosomes ( Supplementary Information and Figure S1 ) . This analysis also provided further information about the progeny and restored “haploid” strains' genetic background: PLB1 was found to be monoallelic in MLST analysis for progeny and restored “haploid” strains but is heterozygous according to the genome data . IGS and TEF1 are also monoallelic in the MLST analysis but biallelic according to the genomic data whereas CAP10 is monoallelic in both MLST and genomic data , as is the MAT locus for which the entire chromosome is monoallelic in all strains . Similar chromosomal abnormalities have previously been described as a common feature in C . gattii and C . neoformans and been suggested as an adaptive mechanism to stresses such as exposure to antifungals [4] , [40]–[44] . In the mating between VGII a ( CBS10090 ) x VGIII α ( NIH312 ) , we isolated a total of 16 progeny . Mitochondrial amplification of the ATP6 gene ( mitochondrial ) revealed that 100% of the progeny exclusively inherited the a mitochondrial genome ( Figure 3B ) . Based on FACS analysis , we found that ∼half ( 7/16 ) were haploid ( 1N ) and the rest ( 9/16 ) were diploid ( 2N ) although this level of analysis cannot distinguish aneuploid isolates ( Figure 3A ) . When each of the progeny was analyzed at eight unlinked MLST loci , including one sex-specific marker ( SXI1/SXI2 ) , 5/16 showed no signs of nuclear exchange ( all α/haploid ) but all five carried an a mitochondrial genome ( Figure 3B ) . This indicates that these five mitochondrial exchange strains are the product of blastospores ( i . e . , monokaryotic yeast budding off of dikaryotic hyphae ) . We also show that two isolates harbor alleles from both nuclear genomes , have uniparental mitochondrial inheritance from the a parent , and also are haploid by FACS analysis , indicating that these two isolates were produced via meiosis , although one of the two shows increased levels of inheritance from the VGIII parent ( Progeny 2 has 7/8 MLST loci from the VGIII parent while P3 has 4/8 MLST loci from the VGIII parent ) . The 9/16 remaining isolates show signs of aneuploidy: they all retain markers with sequences from both parental nuclear genomes and are all 2N or greater than 1N based on FACS analysis ( Figure 3 ) . These nine isolates also show self-fertility and 7/9 have both sex determining alleles ( Figure 3B ) . The remaining two progeny ( 1 and 13 ) show no amplification of the SXI2a allele , however , these are self-fertile α isolates likely exhibiting robust α-α unisexual reproduction . In all of the isolates , the CAP10 locus specifically amplified the VGIII allele , again suggesting either an amplification bias or the loss of this region of chromosome 11 . In the mating between VGIII α ( 97/433 ) x VGII a ( YL4 ) , we isolated a total of 7 progeny as described above . None of the progeny showed signs of self-fertility . Nuclear markers indicated that all progeny except SP130 , which received all tested alleles from the VGIII α parent ( 97/433 ) , were recombinant . Amplification of the ATP6 gene ( encoded by the mitochondrial genome ) revealed that 4/7 of the progeny inherited the a and 3/7 the α mitochondrial genome ( Figure 4A ) indicating non-uniparental inheritance in this outgroup cross . For the crosses with marked strains VGII α ( AIg254 ) x VGIII a ( JF109 ) and VGIII α ( JF101 ) x VGII a ( AIG289 ) three and six viable progeny were isolated . ( Figure 4B&C ) . All nine progeny were recombinant compared to the parental isolates . They also all had inherited their mitochondrial genotype from the a parent . Collectively , our molecular and phenotypic findings indicate that the rate of successful meiosis is low during VGII x VGIII mating with only 2/16 viable progeny being haploid recombinants in a VGII a x VGIII α mating and all of the viable progeny from the VGII α x VGIII a cross being diploid ( 2N ) . Thus , both sets of crosses indicate the presence of a restrictive barrier in meiosis due to cryptic speciation between molecular types VGII and VGIII . Compared to the high germination rate observed between VGIII x VGIII F1 progeny [32] both crosses produced few viable progeny . Although more than 163 spores that could be individually manipulated were produced in VGII x VGIII crosses , the spores did not germinate , suggesting that most progeny were largely inviable , as has been previously reported for sexual crosses between the related species Saccharomyces cerevisiae and S . bayanus [45] , although at lower frequency . Given the involvement of mitochondria in cryptococcal hypervirulence [21] we considered whether mitochondrial genotype is the sole determinant of hypervirulence . If so , then we would anticipate the intracellular proliferation rate and mitochondrial tubularisation pattern of the progeny from these crosses to match that of the mitochondrial-donor parent . Indeed in the outgroup cross R265 x B4546 , all 18 hybrid diploid progeny showed intracellular proliferation rates similar to that of the low-virulence MATa ( mitochondrial donor ) parent B4546 ( Figure 5A ) . In addition , none of these F1 strains were able to trigger extensive mitochondrial tubularisation in response to engulfment by a host macrophage ( Figure 5B ) , suggesting that the replacement of the R265-type mitochondrion with that from B4546 eliminated the hypervirulence trait in these progeny . Because the progeny from these cross were all diploid , we tested whether this hybrid nuclear genotype may be ‘masking’ virulence phenotypes . However , when we “restored six” of the strains to haploidy via repeated rounds of mitotic passage , both mitochondria tubularisation and intracellular proliferation rates remained low ( Figures 5C&D and Figure S2A ) . In contrast to the R265 x B4546 cross , the cross between CBS10090 and NIH312 , in which genotyping indicated all offspring carried mitochondria from the hypervirulent MATa strain CBS10090 , yielded F1 strains showing a wide range of intracellular proliferation rates ( Figure 6A ) . Notably , two haploid recombinant offspring ( Progeny 2 and 3 ) carry mitochondria from the virulent ( CBS10090 ) parent , and yet only one ( Progeny 3 ) shows a high intracellular proliferation rate . Thus a VGII mitochondrial genome , at least alone , is not sufficient to confer hypervirulence in this context . Interestingly , progeny derived from haploid blastospores are isolates in which the nuclear genome is identical to the α parent NIH312 , but the mitochondrial genome has been inherited from the a parent CBS10090 . Such isolates show variable intracellular proliferation rates and tubularisation behaviour , e . g . Progeny 14 presents with IPR similar to the hypervirulent a parent CBS10090 , whereas Progeny 5 , 6 , 15 , and 12 proliferate less well within macrophages . This indicates that additional mechanisms might contribute to hypervirulence . In particular the sex induced silencing pathway becomes activated in blastospore progeny produced during the sexual cycle [46] . Thus , epigenetic processes might also contribute to altering biological properties of blastospore progeny , in addition to the exchange of the mitochondrial genome , leading to modified virulence phenotypes in blastospore progeny that are identical in their nuclear and mitochondrial genomes , yet differ phenotypically . Lastly , we note that recombinant progeny from this cross no longer showed concordance between intracellular proliferation and mitochondrial tubularisation rates upon engulfment , with many strains showing high levels of tubularisation even under control conditions ( Figure 6B and Figure S2B ) . To independently verify these observations , we undertook additional crosses using marked strains , as described above . Crosses between a hypervirulent MATa and a low virulence MATα parent resulted in range of intracellular proliferation rates ( Figure 7A&B ) whereas the reverse cross between a hypervirulent MATα and a low virulence MATa parent only produced progeny with low intracellular proliferative capacity ( Figure 7C ) . Within these outgroup crosses , we also observed misregulated mitochondria ( Figure 7D–F and Figure S2C–E ) . Taken together , the data from these outgroup crosses thus strongly suggest that: For the two ingroup VGII x VGII crosses ( CBS1930 x R265 and LA584 x R265 ) , nuclear markers indicated that all progeny were recombinant , although one ( #37 ) from the cross between CBS1930 and R265 is likely aneuploid , because for one chromosome both parental alleles were amplified . Remarkably , however , for both crosses significant numbers of progeny inherited their mitochondria from the unexpected ( R265 , MATα ) parent: 9/36 for the cross with CBS1930 ( Figure 8A ) and 4/13 for the cross with LA584 ( Figure 8B ) . Thus , it appears that , in contrast to outgroup crosses , crosses within the VGII clade produce a high proportion of viable recombinant progeny but that non-uniparental inheritance of mitochondria ( i . e . in which either parent can donate mitochondria to daughter cells , but not at the same time ) occurs more frequently than anticipated ( 25–30% compared to 5% in previous studies ) . This is analogous to the situation that occurs in atypical diploid-haploid crosses , in which mitochondria are inherited from the MATα parent at a high rate [47] . In contrast to the situation with outgroup crosses , several ingroup crosses resulted in a significant number of progeny that exhibited intracellular proliferation rates that were as high or even higher than the hypervirulent MATα parent ( R265 ) , despite inheriting their mitochondrion from a lower-virulence MATa parent ( Figures 9A&B ) . Conversely , several progeny inherited the mitochondria from the hypervirulent α parent R265 and yet displayed low intracellular proliferation rates . Thus , hypervirulent phenotypes can spread within VGII independently of the mitochondrial genotype . We had anticipated that in these crosses the progeny would inherit mitochondria from a lower virulence MATa parent ( CBS1930 or LA584 ) but , as described above , in fact both crosses demonstrated a higher rate of mitochondrial inheritance from the MATα parent ( 25 to 30% ) , indicating that mitochondrial inheritance is not strictly uniparental in this group . Interestingly , unlike the CBS10090 x NIH312 outgroup cross ( Figure 7 ) , both ingroup crosses produced progeny that remained able to correctly tubularise their mitochondria in response to phagocyte engulfment ( Figures 9C&D and Figure S2F&G ) , which most likely explains the ability of these recombinant progeny to continue to proliferate rapidly within host cells . Increased mitochondrial genome copy number or a higher number of mitochondria can affect mitochondrial inheritance . Similarly , the larger cell size of one parent might lead to an increased cytoplasmic and/or mitochondrial contribution to progeny . For instance , hyper-suppressive RHO mutants of S . cerevisiae exhibit deletions in the mitochondrial genome and are ‘petite’ variants . However , the S . cerevisiae mitochondrial genome replicates faster than the wild type and consequently , when crossed with wild type , all progeny inherit the mutant mitochondrion [48]–[50] . To test whether such a phenomenon may account for the non-uniparental mitochondrial inheritance we observed during in-group crosses , we measured the size of our parental strains under control conditions in vitro and after macrophage passage . However , cell size does not appear to be a contributing factor for changes in virulence ( data not shown ) . In addition , previously published data on mitochondrial DNA copy number [21] showed no increase of mitochondrial genetic information and hence makes it unlikely that a higher copy number from R265 leads to a ‘leak’ from the α parent in those crosses . The surprising result of biparental mitochondrial inheritance in the ingroup crosses ( mitochondria inherited from a parent 70–75% of the time and from the α parent 25–30% of the time ) , in contrast to uniparental inheritance from only the a parent in outgroup crosses , prompted us to further investigate the unexpected phenotypes of progeny from the outgroup cross between CBS10090 and NIH312 . By utilising whole-genome sequence data for the two parental strains and Progeny 5 , which had been made available as part of a larger sequencing project , we found 440 mitochondrial single nucleotide polymorphisms ( SNPs ) that differed between CBS10090 and NIH312 . Progeny 5 shared 320 SNPs with CBS10090 and 120 with NIH312 . Aligning these sites across the 34 kb mitochondrial genome showed that , with the exception of a single SNP , all sites from a single parent formed contiguous blocks ( Figure 10A ) . Thus this pattern of polymorphisms represents very strong evidence for mitochondrial recombination in Progeny 5 from the VGII x VGIII outgroup cross CBS10090 x NIH312 ( Figure 10B ) . In addition , by using three mitochondrial markers to assess the inheritance of mitochondrial DNA in the VGII x VGII crosses , we identified four examples of recombinant mitochondrial genotypes ( Figure 10C ) . These findings support previous evidence for mitochondrial recombination both in C . neoformans [27] , [51] and C . gattii [52] .
In this study we conducted a systematic analysis to test the potential spread of hypervirulence in C . gattii . In crosses between a hypervirulent VGII strain and strains from a different clade , VGIII , it appears that a mitochondrial genotype originating from within the outbreak is necessary , but not sufficient , to confer hypervirulence . Thus simple transmission of a mitochondrial lineage is unlikely to spread outbreak traits to a new population of C . gattii . We generated multiple sets of VGII x VGIII F1 progeny , one from a VGII a x VGIII α mating , and the other from a VGII α x VGIII a mating . The first conclusion from the progeny sets was that germination was very infrequent ( less than 1% , 0/163 spores ) , supporting the hypothesis that these molecular types are cryptic species manifesting strong reproductive barriers . To circumvent this , isolation of progeny had to be done without individual spore dissections ( selecting probable progeny from spore-dense regions ) ; resulting in a need to clearly type all progeny using a multilocus sequence typing ( MLST ) or PCR-RFLP based approaches . Several independent studies have shown that , based on sequence analysis , there is no allelic exchange observed between the four molecular types of C . gattii , leading to a hypothesis that they are independent species within the pathogenic Cryptococcus species complex and adhere to a phylogenetic species concept [11] , [24] , [33] , [34] , [53] . The results presented in this manuscript demonstrate that there are clear reproductive barriers between isolates from two of the molecular types examined , VGII and VGIII . This reproductive barrier is post-zygotic , and supported with examples from both VGII α x VGIII a as well as VGII a x VGIII α crosses . Therefore , these two lineages appear to also adhere to a biological species concept . Specifically , analysis of progeny from such crosses showed predominantly hybrid diploids , followed by mitochondrial exchange strains ( i . e . , blastospores ) , and finally only 1–2/34 ( 3–6% ) haploid recombinants . This post-zygotic barrier parallels studies examining progeny isolated from sexual crosses between the related species Saccharomyces cerevisiae and S . bayanus , whereby viable spores from tetrads were found at a frequency of about 1/10 , 000 [45] . Furthermore , similar post-zygotic barriers between closely related species from within both the Microbotryum and Neurospora genera and AD hybrids in C . neoformans result in low levels of viable progeny [54]–[56] . The finding that a majority of progeny are diploid hybrids is intriguing , as it suggests that , both environmentally and clinically , there may be a potential for inter-molecular type hybrids , such as VGII x VGIII hybrids , to occur . This would parallel several seminal studies examining both C . neoformans var . neoformans/C . neoformans var . grubii ( AD ) and C . neoformans/C . gattii ( BD and AB ) hybrids [41] , [54] , [57]–[62] . These studies show that the hybrids are able to infect hosts , as several are clinical in origin , and our IPR assays , at a minimum , suggest that VGIII x VGII diploid hybrids can be virulent . Mitochondrial inheritance in C . neoformans , as in most eukaryotes , is uniparental [63] , [64] , with progeny inheriting mitochondria from the MATa parent [29] . However , environmental factors such as high temperature and UV irradiation can lead to biparental inheritance and recombination of mitochondrial DNA [65] . Our data indicate that , at least within VGII , inheritance of mitochondria from either parent ( rather than only one ) can occur relatively frequently ( 25–30% ) , even in the absence of such stresses , and that recombination can occur for the mitochondrial genome . Previous studies have shown that while a-α mating leads to uniparental inheritance , regulated by the mating type-specific homeodomain genes SXI1α and SXI2a , α-α mating has biparental inheritance , which enables increased mitochondrial genome recombination [66] . Other recent studies have demonstrated roles for genes not within the mating type locus in the control of uniparental inheritance [67] . Furthermore , in a congenic VGII MATa/MATα strain pair in the R265 background , uniparental inheritance was observed [68] . This supports an interesting hypothesis whereby α-α mating may contribute to the formation of recombinant mitochondrial genomes with higher predispositions for virulence , possibly explaining the hypervirulence among specific VGII genotypes . In addition to demonstrating biparental organelle inheritance in C . gattii , we provide clear genomic evidence for mitochondrial recombination following mating , in support of previous MLST/AFLP data that suggested the occurrence of mitochondrial recombination at the population level [38] , [52] . Together , these data suggest that there may be less stringent control over uniparental mitochondrial inheritance by the mating type locus in the C . gattii VGII lineage as compared to much stricter uniparental mitochondrial inheritance in C . neoformans . Whilst the overall populations of strains still showed a significant correlation between intracellular proliferation and mitochondrial tubularisation ( Figure S2E ) , one striking observation from these crosses was the high frequency of ‘misregulated’ mitochondria , in which mitochondrial tubularisation ( which , in outbreak strains , is a response to the stressful environment of the host cell involving mitochondrial fusion ) occurs even under control growth conditions ( Figure 5 ) . Mitochondrial morphology is regulated by proteins encoded within the nuclear genome and , in S . cerevisiae , fusion events are controlled by a protein complex consisting of Fzo1p , Mgm1p , and Ugo1p [69]–[72] . Thus effective interaction of this machinery might be interrupted by non-compatible mitochondrial and nuclear genomes , reducing the ability of recombinant progeny to regulate virulence traits . In contrast , for crosses between parents within the VGII clade , hypervirulence traits appear to spread easily and are no longer strictly dependent on the presence of a mitochondrial genotype originating from within the outbreak . Thus many , perhaps all , VGII mitochondria are capable of tubularising under host conditions and driving rapid intracellular proliferation , and therefore virulence . This presumably reflects the compatibility of nuclear and mitochondrial ‘cross-talk’ across VGII genotypes , allowing mitochondrial morphology to be correctly regulated in recombinant progeny . This may explain why multiple , distinct outbreaks of disease have all been caused by VGII isolates that differ in their genotype [11] . Our experimental model indicates that hypervirulence in C . gattii is a complex , multigenic trait , requiring regions of the mitochondrial genome and regions of the nuclear genome to confer hypervirulence , which can be attained by a variety of genetic combinations after sexual mating . Thus , there are potentially multiple routes by which such traits could disperse through the C . gattii population , suggesting that surveillance efforts should consider the possibility of independent outbreaks caused by distinct lineages of this pathogen . Our studies also provide phenotypic and molecular evidence that the VGII and VGIII molecular types of C . gattii are distinct species , separated by post-zygotic reproductive barriers . Finally , we provide evidence that mitochondrial inheritance in this species is more complex than currently appreciated , with both biparental mitochondrial inheritance and mitochondrial recombination being observed .
Cryptococcus gattii strains ( Table 1 and Table 2 ) used in this study were cultured in liquid or agar YPD media ( 1% peptone , 1% yeast extract , 2% D- ( + ) -glucose ) for 24 h at 25°C rotating at 20 rpm prior to experiments [73] . Mammalian J774 cells were grown as described previously [73] . For outgroup crosses between VGII a CBS10090 , VGII α R265 , VGIII a B4546 , and VGIII α NIH312 , mating assays were conducted on V8 media ( 5% V8 juice , 0 . 5 g/L KH2PO4 , 4% agar; pH = 5 ) to generate spores for progeny analyses . Isolates were incubated at room temperature in the dark for 2–4 weeks in dry conditions . Fertility was assessed by light microscopy to identify basidiospore formation at the periphery and surface of the mating patch . To collect progeny from the crosses , basidiospores were isolated with a micromanipulator as described previously with the slight modification that due to low germination rate , other suspected resultant colonies were collected from areas where groups of basidiospores were collected [74] , [75] . To summarize , spore-dense regions were collected using a glass Pasteur pipette and spread on a YPD agar plate . In total , the progeny set from the CBS10090 x NIH312 cross yielded 16 collected progeny and the R265 x B4546 cross yielded 18 collected progeny . Progeny from crosses YL4x97/433 , AIg289xJF101 and AIg254xJF109 were established as described above . For AIg289 x JF101 and AIg254 x JF109 , the cell mixture was plated onto YPD supplemented with nourseothricin and G-418 ( both at 100 µg/ml ) . To derive haploid progeny ( SP3-SP8 ) , the diploid progeny were passaged on YPD agar every 48 hours for 24 days ( total 12 passages ) . A single colony from the previous passage was used to initiate the next passage . FACS analysis was performed on five colonies from every passage to determine the ploidy . Serial passaging was stopped when cells from at least one of the five colonies were found to be haploid . Cells from these haploid colonies were used to grow overnight cultures in liquid YPD , which were stored frozen at −80°C in glycerol and subsequently used for MLST and virulence analyses . Crosses between R265 and CBS1930 or LA584 were established on V8 juice agar medium ( pH unadjusted ) . Yeasts were mixed on the plate and examined 2–4 weeks later for the presence of basidia and basidiospore chains . Spores and surrounding parental yeasts were transferred using a gel-loading tip to YPD agar medium . Individual basidiospores were micromanipulated with a dissecting microscope . Genomic DNA from progeny was prepared by disrupting cells in buffer ( 10 mM Tris-HCl [pH 7 . 5] , 10 mM EDTA , 0 . 5 M NaCl , 1% SDS ) +½ volume chloroform with 425–600 µm glass beads , aided by two rounds of vortexing and freezing at −20°C . After centrifugation , the DNA in the supernatant was precipitated with an equal volume of isopropanol . For preparing larger quantities of DNA , a CTAB-based extraction buffer was used on lyophilized cells harvested from 50 ml cultures [76] . For the outgroup crosses sequence data for the MLST alleles of the parental isolates were previously published ( Table S1 ) [11] , [24] . For each F1 progeny isolate , DNA was isolated ( Epicentre ) , and genomic regions were PCR amplified ( Table S2 ) , purified ( ExoSAP-IT , Qiagen ) , and sequenced . Sequences from both forward and reverse strands were assembled and manually edited using Sequencher version 4 . 8 ( Gene Codes Corporation ) . Based on the sequences of the parental strains , alleles of the progeny were assigned to three distinct categories: exclusively from the α parent , exclusively from the a parent , or heterozygous . Heterozygosity at each allele was based on alignments with parental alleles and clear observations of multiple positive nucleotide traces in regions that differ between the two parental sequences , which are indicative of two unique sequences ( i . e . , one from each parent ) being analyzed . Additionally , mitochondrial inheritance was assayed using ATP6 mitochondrial specific forward ( ACTTGCGGCTGAATGATAAAATCTAA ) and reverse ( GTGGAGATGTAATAAAGTGTGTCATG ) primers , whereby the product ( including the 5′ UTR and part of the ORF ) from the VGIII ( a and α ) mitochondrial genome is larger than the product from the VGII ( a and α ) mitochondrial genome [5] , [11] . For progeny from crosses R265 x CBS1930 and R265 x LA584 , polymorphic regions were identified in multilocus sequence typing ( MLST ) data in GenBank and by sequencing fragments of the CBS1930 genome . MLST differences between the strains were used for three PCR-RFLP markers [24] . To identify other polymorphic regions , a small genomic DNA library was constructed from DNA of strain CBS1930 . HindIII restriction fragments were cloned into the HindIII site of plasmid pBluescript . The ends of the inserts were sequenced , and the sequence compared by BLASTn to the R265 strain genome database at the Broad Institute [77] . Either single nucleotide polymorphisms ( SNPs ) affecting restriction enzymes sites or with multiple differences between the two strains were used for the design of oligonucleotides primers for PCR amplification of alleles from either parent . PCR reactions , digested with restriction enzyme where necessary , were resolved on 1× TAE agarose gels . Primer sequences and details about the polymorphisms are in Table 3 . Sections of the mitochondrial genome were amplified and sequenced from R265 , CBS1930 , and LA584 to identify a polymorphism . Analysis of nuclear and mitochondrial markers was conducted to test if progeny were recombinant was conducted . The strains from the cross between R265 and CBS1930 were assessed for 16 genetic markers . One was the mating type phenotype and 14 were PCR markers that amplified polymorphic parts of the nuclear genome . The markers are located on eight of the 14 chromosomes of C . gattii [77] . For the mitochondrial genome , the COX1 and COB1 genes were amplified and sequenced . A single nucleotide polymorphism was identified in intron 2 of COB1 ( submitted as GenBank accessions JX486912 and JX486913 ) . Subsequently , the CBS1930 strain was subject to Illumina genome sequencing , and the mitochondrial genome analyzed for other differences between this strain and R265 . Two other regions were used to track the inheritance of the mitochondrial genome in the VGII x VGII progeny . Cells were processed for flow cytometry as described previously , with slight modifications [54] , [78] . Briefly , cells were harvested from YPD medium , washed once in phosphate-buffered saline ( PBS ) buffer , and fixed in 1 ml of 70% ethanol overnight at 4°C . Fixed cells were washed once with 1 ml of NS buffer ( 10 mM Tris-HCl ( pH 7 . 5 ) , 250 mM sucrose , 1 mM EDTA ( pH 8 . 0 ) , 1 mM MgCl2 , 0 . 1 mM CaCl2 , 0 . 1 mM ZnCl2 ) and then stained with propidium iodide ( 12 . 5 mg/ml ) in 0 . 2 ml of NS buffer containing RNaseA ( 1 mg/ml ) at 4°C for 16 h . Next , 0 . 5 ml of stained cells were diluted into 0 . 5 ml of 50 mM Tris-HCl ( pH 8 . 0 ) . Flow cytometry was performed on 10 , 000 cells and analyzed on the FL1 channel with a Becton-Dickinson FACScan ( Duke University Medical Center Flow Cytometry Core Facility ) . Filamentation assays were conducted on V8 media ( pH = 5 ) and filamentation agar [79] . Isolates were incubated at room temperature in the dark for 2–4 weeks in dry conditions . Filamentation was assessed by light microscopic examination for hyphae formation at the periphery and surface of the incubated patches . All assays were conducted on both media types . If there were no signs of filamentation after a four-week period , isolates were scored as having no self-filamentation phenotype . Macrophages were infected with yeast cells and intracellular proliferation monitored as previously described [80] . Cryptococcal mitochondrial morphology was determined as described previously [11] . In brief , to determine the intracellular proliferation rate ( IPR ) of individual strains following phagocytosis , J774 macrophage cells were exposed to cryptococcal cells that were opsonized with 18B7 antibody ( a kind gift from Arturo Casadevall ) for 2 hr as described previously [73] . Each well was washed with PBS in quadruplicate to remove as many extracellular yeast cells as possible and 1 ml of fresh serum-free DMEM was then added . For time point T = 0 , the 1 ml of DMEM was discarded and 200 µl of sterile dH2O was added into wells to lyse macrophage cells . After 30 minutes , the intracellular yeast were released and collected . Another 200 µl dH2O was added to each well to collect the remaining yeast cells . The intracellular yeast were then counted with a haemocytometer . For the subsequent five time points ( T = 18 hrs , T = 24 hrs , T = 48 hrs , T = 72 hrs ) , intracellular cryptococcal cells were collected and counted . For each strain tested , the time course was repeated at least three independent times , using different batches of macrophages . The IPR value was calculated by dividing the maximum intracellular yeast number by the initial intracellular yeast number at T = 0 . We confirmed that Trypan Blue stains 100% of the cryptococcal cells in a heat-killed culture , but only approximately 5% of cells from a standard overnight culture . The mitochondrial morphology assays were conducted in a similar way to those in previous studies , with modifications [21] . C . gattii cells were grown overnight at 37°C in DMEM untreated or isolated from macrophages 24 hr after infection . The cells were harvested , washed with PBS twice and re-suspended in PBS containing the Mito-Tracker Red CMXRos ( Invitrogen ) at a final concentration of 20 nM . Cells were incubated for 15 min at 37°C . After staining , cells were washed three times and re-suspended in PBS . For each condition , more than 100 yeast cells per replicate for each of the tested strains were chosen randomly and analysed . To quantify different mitochondrial morphologies , images were collected using a Zeiss Axiovert 135 TV microscope with a 100× oil immersion Plan-Neofluar objective or a Nikon Eclipse Ti Plan Apo VC 60× oil immersion objective . Both fluorescence images and phase contrast images were collected simultaneously . Images were captured with identical settings on a QIcam Fast 1394 camera using the QCapture Pro51 version 5 . 1 . 1 software . All images were processed identically in ImageJ and mitochondrial morphologies were analysed and counted blindly [11] . IPR and tubularisation data were analysed for statistically significant differences using one-way ANOVA analysis with multiple comparisons by Tukey Honestly Significant Difference ( HSD ) posthoc test . A p-value of <0 . 05 after controlling for multiplicity was considered to be statistically significant . Genomic DNA from C . gattii strains NIH312 , CBS10090 , and progeny 5 from the cross between NIH312 and CBS10090 was isolated with the EpiCentre MasterPure Yeast DNA Purification Kit according to a modified version of the instruction manual . Briefly , the strains were grown in liquid YPD media for 24 h at 25°C rotating at 20 rpm . Cells from 3 ml of culture were harvested by centrifugation at 17 , 000× g for 5 minutes . Cells were lysed in 300 µl of Yeast Cell Lysis solution by mechanical disruption with 0 . 1 mm silica spheres ( FastPrep Lysing Matrix , MP Biomedicals ) twice for 30 seconds at 6 , 800 rpm in a Precellys24 and incubation at 65°C for 15 minutes . Samples were cooled down on ice for 5 minutes and proteins removed by vortexing with 150 µl of MPC Protein Precipitation Reagent and following centrifugation for 10 minutes at 17 , 000× g . DNA was recovered with 500 µl isopropanol and centrifugation at 17 , 000× g for 10 minutes . DNA was purified by RNase A treatment for 60 minutes at 37°C followed by phenol:chloroform extraction and ethanol precipitation . DNA yield and quality was determined by spectrophotometry . 2 µg of genomic DNA were used for library preparation: DNA was fragmented to 150–500 bp using Covaris shearing and processed with the TruSeq DNA Sample Prep Kit ( Illumina ) according to instructions , purification steps were performed with Agencourt AMPureXP magnetic particles ( Beckham Coulter ) on a magnetic stand ( AmBio ) . Whole genomes were sequenced on an Illumina HiSeq2000 at the MRC Clinical Science Centre , Imperial College London ( UK ) . Alignment and SNP calling parameters were initially optimized . The nuclear and mitochondrial genome sequences and feature files for C . gattii isolate R265 ( VGII ) were downloaded from http://www . broadinstitute . org/ ( GenBank project accession number AAFP01000000 ) . Illumina reads were aligned to the genome sequence using Burrows-Wheeler Aligner ( BWA ) v0 . 5 . 9 [81] with default parameters and converted to pileup format using Samtools v . 0 . 1 . 18 [82] . To act as a control for sequencing , alignment and SNP calling , we resequenced the reference strain R265 . We used a False Discovery ( FDR ) approach [83] to test our SNP-calling method , which we set at a minimum required depth of four reads , with 90% disagreeing from the reference base and agreeing with each other . First , we randomly modified 63 , 193 and 698 , 535 nucleotides within the reference sequence , corresponding with the maximum number of SNPs identified within the VGII group and within any of our C . gattii isolates , respectively . We aligned the reads of R265 to these two altered genome sequences and called SNPs using our chosen parameters . We identified 99 . 32% and 99 . 26% of true positives , whilst only calling 3553 ( 5 . 6% ) and 3363 ( 0 . 48% ) false positives or genuine discrepancies with the reference sequence respectively . For analysis we used all SNPs that were covered by ≥4 reads in all isolates , leaving a total 740 mitochondrial sites . To ensure that these sites could not have been heterogeneous in the progeny sequence we examined the allele frequencies at each variant site and found that each site had greater than 95% agreement with the called SNP . We simulated 100 bp reads with 0 . 01% uniform error for 200× coverage from each parental mitochondrial genotype and used them in a single combined BWA alignment and SNPs calling protocol as before . We detected only seven sites that show shared differences from the reference sequence of R265 . The other SNPs were not detected because they did not reach the 90% agreement threshold . Two to four million 100 bp Illumina paired end reads were assembled with velvet ( version 1 . 2 . 08 ) [84] . Redundant assembly runs with varying k-mer lengths were performed for each strain , and each time resulted in identical circular contigs , which only differed in lengths of overlapping ends . Supercontigs of C . gattii R265 were obtained from Broad Institute ( http://www . broadinstitute . org ) , and grouped in chromosomal context by alignment to C . gattii WM276 chromosomes [77] with MUMmer ( version 3 . 23 ) [85] , [86] . The resulting tiled R265 chromosomes served as reference in downstream analyses . Ploidy of EJB and SP strains was assessed by read coverage and allele frequencies at variant sites . After mapping with Bowtie 2 ( version 2 . 1 . 0 ) [87] , read coverages were calculated and plotted with CNV-seq [88] . C . gattii WM276 was used as mapping genome , a theoretical VGII x VGIII diploid of equally pooled R265 , B4546 reads as reference , and R265 x B4546 progeny ( Progeny and SP strains ) reads as testing samples , respectively . Allele frequencies of variant sites in Progeny and SP progeny were calculated after Bowtie 2 read mapping to R265 chromosomes , and variant calling with samtools/bcftools from the samtools package ( version 0 . 1 . 18 ( r982:295 ) ) [82] Allele frequencies were extracted from DP4 fields of VCF output , and nucleotide variant ratios calculated for each position by division of reads depth/number of variant bases . A separate mapping of R265 reads was performed as control , and observed positions removed as background noise from Progeny and SP variant calls . Additional details on the ploidy analyses are available in the supplementary information ( Text S1 ) . To assess mitochondrial recombination , genomes of CBS10090 and NIH312 were compared to CBS10090 x NIH312 progeny strain Progeny 5 . Sequences were aligned , and corresponding regions visualized with Progressive Mauve ( version 2 . 3 . 1 ) [89] . Additional details on the analysis of mitochondrial genomes are available in the supplementary information ( Text S1 ) . | How infections spread within the human population is an important question in forecasting potential epidemics . One way to investigate potential mechanisms is to test experimentally whether combinations of genes that confer high virulence are able to spread to less-virulent lineages . Here , we address this question in a fungal pathogen that is causing an outbreak of meningitis in healthy humans in Canada and the Pacific Northwest . We demonstrate that virulence traits are easily transmitted between closely related pathogenic strains , but are more difficult to transmit to more distant lineages . In addition , we show that a paradigm of organelle inheritance , namely that mitochondria are inherited uniparentally from the a mating type , is altered in the R265α outbreak strain such that it transmits its mitochondrial genome to 25–30% of its progeny . This biparental inheritance likely contributes to increased mitochondrial recombination . Taken together , our data suggest that virulence traits may be relatively mobile within this species and that current models of mitochondrial inheritance may require revising . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Transmission of Hypervirulence Traits via Sexual Reproduction within and between Lineages of the Human Fungal Pathogen Cryptococcus gattii |
Chagas disease ( CD ) affects over six million people and is a leading cause of cardiomyopathy in Latin America . Given recent migration trends , there is a large population at risk in the United States ( US ) . Early stage cardiac involvement from CD usually presents with conduction abnormalities on electrocardiogram ( ECG ) including right bundle branch block ( RBBB ) , left anterior or posterior fascicular block ( LAFB or LPFB , respectively ) , and rarely , left bundle branch block ( LBBB ) . Identification of disease at this stage may lead to early treatment and potentially delay the progression to impaired systolic function . All ECGs performed in a Los Angeles County hospital and clinic system were screened for the presence of RBBB , LAFB , LPFB , or LBBB . Patients were contacted and enrolled in the study if they had previously resided in Latin America for at least 12 months and had no history of cardiac disease . Enzyme-linked immunosorbent assay ( ELISA ) and immunofluorescence assay ( IFA ) tests were utilized to screen for Trypanosoma cruzi seropositivity . A total of 327 consecutive patients were screened for CD from January 2007 to December 2010 . The mean age was 46 . 3 years and the mean length of stay in the US was 21 . 2 years . Conduction abnormalities were as follows: RBBB 40 . 4% , LAFB 40 . 1% , LPFB 2 . 8% , LBBB 5 . 5% , RBBB and LAFB 8 . 6% , and RBBB and LPFB 2 . 8% . Seventeen patients were positive by both ELISA and IFA ( 5 . 2% ) . The highest prevalence rate was among those with RBBB and LAFB ( 17 . 9% ) . There is a significant prevalence of CD in Latin American immigrants residing in Los Angeles with conduction abnormalities on ECG . Clinicians should consider evaluating all Latin American immigrant patients with unexplained conduction disease for CD .
Chagas disease ( CD ) , caused by the protozoan Trypanosoma cruzi , is a slow-progressing , multi-organ disease endemic to Latin America . There are an estimated 6 million infected individuals worldwide . [1 , 2] CD has an acute and chronic phase , with the chronic phase beginning 4–8 weeks after the initial infection . [3] The chronic phase begins in an asymptomatic indeterminate form characterized by seropositivity for antibodies against T . cruzi , a normal electrocardiogram ( ECG ) , and a normal chest radiograph . Without treatment , at least 30–40% of patients with the indeterminate form will develop an advanced or determinate form 10–30 years after the initial infection . [3 , 4] The advanced chronic form of CD can lead to irreversible cardiac damage resulting in conduction disease , apical aneurysms , cardiomyopathy , and sudden cardiac death . [4] CD is traditionally associated with endemic regions in Latin America . However , given migration trends , there has been increasing recognition of populations with CD in Europe and the United States . A recent meta-analysis of European studies , which in aggregate screened 10 , 000 Latin American immigrants , found a CD prevalence of 4 . 2% . [5] Another study estimates 300 , 000 cases of CD in the US , contributing to 30–45 , 000 cases of cardiomyopathy . [1] Between 2007 and 2013 , 1908 cases of CD were identified in the blood donation system . [6] In a study of blood samples in Los Angeles , 1 in 1 , 993 were positive for T . cruzi antibodies . [7] Nonetheless , an overwhelming majority of CD patients in the US are undiagnosed and untreated . [6 , 8] Conduction disorders are characteristic of chronic determinate Chagas disease , and are often the initial presenting finding . A study in Bolivia found ECG abnormalities in 46% of seropositive children , the most frequent being incomplete right bundle branch block ( RBBB ) . [9] Another study in Mexico found that ECG abnormalities including RBBB were significantly higher among seropositive versus seronegative individuals . [10] In a sample of 1 , 389 people in a rural community of Brazil with a T . cruzi prevalence of 6 . 6% , ECG abnormalities were observed in 43 . 5% of seropositive compared with 18 . 3% of seronegative individuals . [11] Further , ECG abnormalities can help identify patients who are at higher risk of developing impaired systolic function . The presence of ECG abnormalities at baseline was a significant predictor of decrease in left ventricular ejection fraction ( LVEF ) after 17 months of follow-up in a cohort of Brazilian patients . [12] Conduction abnormalities and cardiomyopathy are also strongly associated with CD in Latin American immigrants in the United States and Europe . In our center in Los Angeles , among adult Latin American patients with nonischemic cardiomyopathy , defined as an LVEF <40% , we found a CD prevalence of 19 . 2% . [13] Another study in New York identified five seropositive cases among 39 immigrants from CD-endemic countries with dilated cardiomyopathy , a prevalence of 13% . [14] Among a sample of 17 T . cruzi-positive blood donors in southeast Texas , 7 ( 41% ) exhibited evidence of cardiomyopathy on electrocardiograph . [15] In Spain , an investigation of 485 T . cruzi-positive immigrants , of whom 459 ( 94 . 6% ) were Bolivian , determined 31 . 5% had at least one ECG abnormality . [16] The purpose of this study is to assess the prevalence of CD in a population of Latin American immigrants with conduction abnormalities on electrocardiogram in a Los Angeles County Hospital .
We computed frequencies and proportions for categorical variables , and means and standard deviations for continuous variables . Chi-square tests for independence or Fisher’s exact tests , as appropriate , were used to detect associations between categorical variables , and t-tests were employed for continuous variables . All p values are two-sided , with p < 0 . 05 considered significant for all analyses . Analyses were conducted with SPSS software , version 23 ( SPSS Inc . , Chicago IL ) . The study was approved by the Institutional Review Board at Olive View-UCLA Medical Center . All participants provided written informed consent prior to participating . There was no compensation for participation .
Study participants had a mean age of 46 . 3±10 . 8 years and had resided in the U . S . for a mean of 21 . 3±10 . 7 years ( Table 1 ) . Countries of origin for the study sample were Mexico ( n = 197 , 60 . 2% ) , El Salvador ( n = 70 , 21 . 4% ) , Guatemala ( n = 31 , 9 . 5% ) and other ( n = 29 , 8 . 9%: Honduras 6 , Peru 6 , Nicaragua 5 , Argentina 5 , Costa Rica 2 , Colombia 2 , Bolivia 2 , and Chile 1 ) . Conduction abnormalities among the study group were as follows: RBBB 40 . 4% , LAFB 40 . 1% , LPFB 2 . 8% , LBBB 5 . 5% , RBBB and LAFB 8 . 6% , and RBBB and LPFB 2 . 8% ( Table 2 ) . Seventeen patients were positive for T . cruzi by both IFA and ELISA , resulting in an overall prevalence rate of 5 . 2% in this cohort of patients with unexplained conduction disease . These patients had not been previously diagnosed and were unaware they had CD . In the seropositive group , the mean age was 50 . 8±10 . 7 years with a mean time of residence in country of origin of 28 . 1±10 . 0 years . The difference in mean ages ( 4 . 8 years ) between the seropositive and seronegative group was not statistically significant at the p<0 . 05 level ( p = 0 . 08 ) . A much smaller proportion of seropositive patients ( n = 5 , 29 . 4% ) were male , compared with the seronegative group ( n = 170 , 54 . 8% ) , and this difference was significant ( p = 0 . 048 ) . The countries of origin of the seropositive patients were as follows ( prevalence within subgroup in parentheses ) : El Salvador 8 ( 11 . 4% ) , Mexico 5 ( 2 . 5% ) , Guatemala 2 ( 6 . 5% ) , and other 2 ( 6 . 9% ) ( Table 1 ) . There was substantial variation between countries; the prevalence was significantly lower for Mexicans yet higher for Salvadorans ( p = 0 . 001 ) . We found the following conduction abnormalities within the seropositive group: RBBB ( n = 7 , 41 . 2% ) , LAFB ( n = 5 , 29 . 4% ) , and RBBB in conjunction with LAFB ( n = 5 , 29 . 4% ) ( Fig 1 , Table 2 ) . No positive patients had LBBB , LPFB , or RBBB and LPFB . We calculated CD prevalence according to each type of conduction abnormality . For RBBB , 7/132 patients ( 5 . 3% ) were seropositive , for LAFB , 5/131 ( 3 . 8% ) , and for RBBB/LAFB 5/28 ( 17 . 9% ) . The risk for positive CD diagnosis in patients with both RBBB and LAFB , compared to other conduction abnormalities in the sample , was five times greater ( OR = 5 . 2 , CI = 1 . 7–16 . 0 , p = 0 . 002 ) .
We did not account for potentially confounding factors such as age , diabetes mellitus , or hypertension in our analyses . The subgroup of seropositive patients was small , creating wide confidence intervals in the calculation of risk factors . Exclusion of patients with underlying cardiac disease could possibly lead to an underestimation of prevalence of CD . This study is based on a sample of patients from a Los Angeles County public hospital system; the results may not be generalizable to other locations . | Chagas disease ( CD ) affects an estimated 300 , 000 people in the United States , but warning signs for the disease have not been closely studied . CD is usually acquired in Latin America , and can remain in the body for years or decades without producing any symptoms . However , in about 30% of patients , it can eventually result in heart failure and death . The electrocardiogram can detect potential heart problems before patients begin to feel symptoms , providing an early warning . If patients with CD receive monitoring and treatment in time , it may prevent the development of more serious heart problems . We checked for the presence of CD in a sample of 327 patients with abnormal electrocardiogram readings , all of whom had resided in Latin America for at least 12 months . Seventeen patients , or 5 . 2% of the total sample , were positive for CD . Our study discusses the association of different electrocardiogram readings with CD in the United States , and explores variations based on patients’ gender and country of origin . The electrocardiogram can be a valuable tool for detecting and measuring the progression of CD in patients from Latin America so that proper treatment can be offered . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"united",
"states",
"cardiomyopathies",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"biochemical",
"analysis",
"ethnicities",
"north",
"ame... | 2017 | Prevalence of Chagas Disease in a U.S. Population of Latin American Immigrants with Conduction Abnormalities on Electrocardiogram |
Experiments show that spike-triggered stimulation performed with Bidirectional Brain-Computer-Interfaces ( BBCI ) can artificially strengthen connections between separate neural sites in motor cortex ( MC ) . When spikes from a neuron recorded at one MC site trigger stimuli at a second target site after a fixed delay , the connections between sites eventually strengthen . It was also found that effective spike-stimulus delays are consistent with experimentally derived spike-timing-dependent plasticity ( STDP ) rules , suggesting that STDP is key to drive these changes . However , the impact of STDP at the level of circuits , and the mechanisms governing its modification with neural implants remain poorly understood . The present work describes a recurrent neural network model with probabilistic spiking mechanisms and plastic synapses capable of capturing both neural and synaptic activity statistics relevant to BBCI conditioning protocols . Our model successfully reproduces key experimental results , both established and new , and offers mechanistic insights into spike-triggered conditioning . Using analytical calculations and numerical simulations , we derive optimal operational regimes for BBCIs , and formulate predictions concerning the efficacy of spike-triggered conditioning in different regimes of cortical activity .
The cerebral cortex contains interacting neurons whose functional connections are modified through repeated patterns of activation . For example , motor and somatosensory cortices are typically organized into somatotopic regions in which localized neural populations are associated with muscles or receptive fields and show varied levels of correlated activity ( e . g . [1–5] ) . Functional relationships between such neural populations are known to change over time , reinforcing relevant pathways [6–9] . These changes are the result of plasticity mechanisms acting on myriad synaptic connections between cortical neurons . Most of them are relatively weak but can potentiate under the right conditions . However , it is not always clear what such conditions might be , or how one can interact with them for experimental or clinical purposes . Unanswered questions include the way local synaptic plasticity rules lead to stable , emergent functional connections , and the role of neural activity—and its statistics—in shaping such connections . While recent and ongoing work elucidates various plasticity mechanisms at the level of individual synapses , it is still unknown how these combine to shape the recurrently connected circuits that support population-level neural computations . Bidirectional brain-computer interfaces ( BBCI ) capable of closed-loop recording and stimulation have enabled targeted conditioning experiments that probe these issues . In a seminal experiment [10] , a BBCI called the Neurochip recorded action potentials of a neuron at one site in motor cortex ( MC ) and delivered spike-triggered stimuli at another for prolonged times in freely behaving macaque monkeys . This conditioning was able to increase functional connectivity from neurons in the recorded site to the ones in the stimulated site ( c . f . [11] ) , as measured by electromyogram ( EMG ) of muscle activation evoked by intracortical microstimulation ( ICMS ) in MC . Importantly , the relative strength of induced changes showed a strong dependence on the spike-stimulus delay , consistent with experimentally derived excitatory spike-timing-dependent plasticity ( STDP ) time windows [12–14] . The effects of this protocol were apparent after about a day of ongoing conditioning , and lasted for several days afterwards . Similar spike-triggered stimulation showed that corticospinal connections could increase or decrease , depending on whether the postsynaptic cells were stimulated after or before arrival of the presynaptic impulses [15] . This BBCI protocol has potential clinical uses for recovery after injuries and scientific utility for probing information processing and learning in neural circuits . The observations outlined above suggest that STDP is involved in shaping neural connections by normal activity during free behavior , and is the central mechanism behind the success of spike-triggered conditioning . However , this could not be verified directly as current experiments only measure functional relationships between cortical sites . Furthermore , interactions between BBCI signals and neural activity in recurrent networks are still poorly understood , and it remains unclear how BBCI protocols can be scaled up , made more efficient , and optimized for different experimental paradigms . For example , during spike-triggered stimulation , the spikes from a single unit are used to trigger stimulation of a neural population . While STDP can explain how the directly targeted synapses may be affected ( i . e . from the origin of the recorded spikes to the stimulated population ) , the observed functional changes must rely on a broader scope of plastic changes involving other neurons that are functionally related to the recorded ones . What are the relevant network mechanisms that govern these population-level changes ? How can a BBCI make use of population activity to trigger optimal stimulation ? Here we advance a modeling platform capable of capturing the effects of BBCI on recurrent , plastic neuronal networks . Our goal is to use the simplest dynamical assumptions in a “bottom-up” approach , motivated by neuronal and synaptic physiology , that enable the reproduction of key experimental findings from [10] at the functional level in MC and related muscles . In turn , we seek to use this model to provide insights into plasticity mechanisms in recurrent MC circuits ( and other cortical regions ) that are not readily accessible experimentally , as well as establish a theoretical framework upon which future BBCI protocols can be developed . We build on a well-established body of work that enables analytical estimates of synaptic changes based on network statistics [16–22] and compare theoretical results with experiments and numerical simulations of a probabilistic spiking network . In our model , every neuron is excitatory—the modulatory role of inhibition in MC is instead represented implicitly by non-homogeneous probabilistic activation rates . While inhibition likely plays an important role in cortical dynamics , we consider results from our exclusive use of excitation to be a significant finding , suggesting that a few key mechanisms can account for a wide range of experimental results . Using data from previous work as well as from novel experiments , we calibrate STDP synaptic dynamics and activity correlation timescales to those typically found in MC neural populations . The result is a spiking model with multiplicative excitatory STDP and stable connectivity dynamics which can reproduce three key experimental findings: Furthermore , we make the following novel findings: Together , these results provide quantifiable experimental predictions . They arise from a theoretical framework that is easily scalable and serves as a potential testbed for next-generation applications of BBCIs . We discuss ways to use this framework in state-dependent conditioning protocols .
When neurons in the network’s three groups a , b , c are subject to external commands ν ( t ) = ( νa ( t ) , νb ( t ) , νc ( t ) ) with stationary statistics , their averaged connectivity J ¯ ( t ) evolves toward an equilibrium J ¯ * that reflects these inputs’ correlations ( c . f . [24 , 25] ) , although individual synapses may continue to fluctuate . This has been observed in a number of theoretical studies ( see e . g . [18 , 23] ) and is consistent with the formation and dissociation of muscle assemblies in MC due to complex movements that are regularly performed [8] . The mean synaptic equilibrium J ¯ * strongly depends on the external inputs ν ( t ) ’s correlation structure C ^ ( u ) ( see Fig 2A ) . Indeed , a narrow peak near the origin for correlations within groups , as is the case for the periodic external rates shown in Fig 1B , along with the absence of such peaks for cross-group correlations , contribute to strengthening synapses within groups and weakening those across groups . Under such conditions , what will be the impact of spike-triggered stimulation ? Fig 2B shows the evolution of synaptic averages J ¯ α β ( t ) , analytically computed ( see Methods ) for a system initiated at the synaptic equilibrium associated with external rates ν ( t ) from Fig 1B . The inset of Fig 2B shows the evolution of individual synapses from group a to group b from full network simulations . At 15 hours , the spike-triggered stimulation protocol is turned “on” , with a set delay d† = 20 milliseconds , and synapses start changing . In ∼10 hours they reach a new equilibrium which differs from the initial one in a few striking ways , as seen in Fig 2C and 2D , where normalized differences ( J ¯ † - J ¯ ) / J m a x are plotted for all pre- and post-group combinations . First , as expected and in accordance with experiments [10] , the mean strength of synapses from group a ( recorded site ) to group b ( stimulated site ) are considerably strengthened ( by about 80% ) . As described in more detail below , this massive potentiation relies on two ingredients: correlated activity within group a and an appropriate choice of stimulation delay d† . Perhaps more surprising are collateral changes in other synapses , although they are of lesser magnitude . While this was previously unreported , it is consistent with unpublished data from the original spike-triggered stimulation experiment [10] . It is unclear how many of these changes are due to the particular external rate statistics and other model parameters; we return to this question below when realistic activity statistics are considered . We also show that spike-triggered stimulation induces novel correlation structures due to synaptic changes as illustrated in Fig 2A , which plots the correlation functions C ^ ( u ) , C ( u ) , C† ( u ) and C J † ( u ) . Here , C J † ( u ) denotes the correlations one observes under baseline activity ( i . e . without ongoing stimulation ) but with the equilibrium connectivity J ¯ † * obtained after prolonged spike-triggered stimulation , i . e . , at the end of spike-triggered stimulation . It is clear that every interaction involving group b is considerably changed , most strongly that with group a , which includes the neuron used to trigger stimulation . More surprising is the increased cross-correlation of group b with itself , even though connectivity within group b is not explicitly potentiated by conditioning . In fact , it is slightly depressed ( Fig 2D ) . This happens because connections from group a to group b are considerably enhanced , which causes the mean firing rate of group b to grow and its correlations to increase . Later , we explore similar collateral changes that occur because of multi-synaptic interactions . We see in the next section how these correlations translate into functional changes in the network . Finally , Fig 2B shows a crucial feature of our model: the timescale for the convergence from the normal to the artificial equilibrium is different from that of the decay back to normal equilibrium after the end of spike-triggered stimulation . In the conditioning experiments from [10 , 15] , the effect of spike-triggered stimulation was seen after about 5–24 hours of conditioning , while the changes decayed after 1 to 7 days . With the simplified external drives producing a reasonable mean firing rate of about 10Hz for individual cells , an STDP learning rate of η = 10−8 was adequate to capture the two timescales of synaptic changes . Thus , the simple excitatory STDP mechanisms in our model give rise to distinct timescales for increases and decay of synaptic connectivity strength produced by spike-triggered conditioning , in agreement with experimental observations . The emergence of distinct timescales was previously reported and studied in related modelling contexts [30 , 31] . It should be noted that a number of parameters are shown to affect the magnitude of timescale separation , such as types of synaptic delays , weight dependence of the STDP rule , the firing rate of ongoing baseline activity , etc . We reiterate that many of these parameters are not well resolved experimentally for macaque MC , and that we have made simplifying choices ( see Methods ) that can be adapted to new experimental data . Nevertheless , we expect our model to robustly produce distinct synaptic timescales that can be fitted by a single time-scaling parameter . This does not contradict the findings from human psychophysical studies that feedback error-driven motor adaptation may involve two or more different and independent parallel processes with different temporal dynamics for learning and decay of the motor skill [35] . Nevertheless , the cellular mechanisms in our model may have some relation to the different timescales proposed to underlie motor adaptation at the system level [35] . Such relationships could be further investigated by direct experimentation and appropriate simulations . In summary , our model satisfies the first experimental observation from [10] we set out to reproduce ( point a . in Introduction ) . Indeed , we find that two distinct timescales of synaptic changes ( during and after conditioning ) are an emergent property of our model , and tuning a single parameter is sufficient to fit the rates observed in experiments . Changes in correlations due to spike-triggered conditioning indicate that there is an activity-relevant effect of induced synaptic changes , which is measurable from spiking statistics ( see Fig 2A ) . We now show how this is directly observable in evoked activity patterns that are consistent with intra-cortical microstimulation ( ICMS ) protocols employed in experiments . In [10] , connectivity changes were inferred using ICMS and electromyogram ( EMG ) recordings of the monkey’s wrist muscles , as well as evoked isometric torques . To summarize , a train of ICMS stimuli lasting 50 ms was delivered to each MC site; simultaneously , EMG activity in three target muscles were recorded . The average EMG responses for repeated trials were documented for each of three MC sites ( i . e . group a , b and c ) before and after spike-triggered conditioning . The experiment showed that prior to conditioning , ICMS stimulation of any MC site elicited well-resolved average EMG responses , largest in one muscle but not the two others . After conditioning , ICMS stimulation of the recording site ( group a ) not only elicited an EMG response in its associated muscle , but also in that of stimulated site ( group b ) . While it was conjectured that synaptic changes in MC were responsible for the changes , this could not be verified directly . Our model suggests that synaptic changes can indeed occur in MC-like networks , but it remains unclear if such changes can lead to the experimentally observed motor output changes . We address this by simulating EMG responses of our model , before and after spike-triggered conditioning ( † ) . Fig 3 shows a simulated ICMS protocol before ( panel A ) and after ( panel B ) spike-triggered stimulation conditioning . For each case , synaptic matrices are chosen from full network simulations and fixed ( STDP is turned off ) , as shown in the top row of Fig 3 ( reproduced from Fig 2C ) . To mimic the ICMS stimulus , we add a square-pulse input rate of 100 Hz lasting 50 ms to the external rate να ( t ) of a target group . An example of the spiking output of our network for α = a is shown in the top row of Fig 3 where the solid black bar below the graph shows the stimulus duration . Next , we filter the spike output of all neurons within a group using the synaptic filter ε ( t ) described in Methods , and add them to obtain population activity time-courses . Finally , we take these summed profiles and pass them through a sigmoidal non-linearity— ( 1 + exp−a ( x − b ) ) −1 where x is the filtered activity—meant to represent the transformation of neural activity to EMG signals . Here , we assume that the hypothetical motoneurons whose target muscle EMG is recorded receive inputs only from a single neural group and that network interactions are responsible for cross-group muscle activation . We label our modelled motor output EMG measurements by Mα , α ∈ {a , b , c} . We choose the nonlinearity parameters a = 2 . 5 and b = 5 to qualitatively reproduce the EMG responses seen in the experiment before spike-triggered conditioning: namely , well-resolved EMG responses Mα are observed only when the relevant MC group α is stimulated . The bottom row of Fig 3 shows Mα responses of each group , averaged over 15 trials each , when ICMS stimulation is delivered to a single group at a time . Panel A shows little cross-group response to ICMS stimulation before spike-triggered stimulation conditioning . However , after conditioning , stimulation of group a evokes an emergent response in the muscle activated from group b ( see circles in Fig 3 ) , as well as a small increase for group c . These features were both present in the original experiments ( see Figure 2 in [10] ) and are consistent with the synaptic strengths across groups before and after conditioning . In addition to EMG , the authors of [10] also measured the effects of ICMS using a manipulandum that recorded isometric torques produced by evoked wrist motion . In our model , the newly evoked EMG responses , after conditioning , agree with the observation that torques evoked from the recorded site ( group a ) typically changed toward those previously evoked from the stimulated site ( group b ) . As such , from now on we equate an increase in mean synaptic strength J ¯ α β between groups to an increase in functional connectivity . We conclude that our model satisfies the second experimental observation from [10] we set out to reproduce ( point b . in Introduction ) . That is , a simple interpretation of evoked network activity—a filtered output of distinct neural group spiking activity—is consistent with the functional changes in muscle activation in ICMS protocols observed before and after conditioning . Up to now , we used toy activation profiles ν ( t ) in the form of truncated sinusoidal bumps to drive neural activity ( Fig 1B ) . In this section , we modify our simple model to incorporate experimentally observed cross correlation functions , whenever possible , in an effort to eliminate artificial activation commands and capture more realistic regimes . As a result , we no longer rely on numerical simulations of spiking activity , but rather on analytically derived averaged quantities to explore a wide range of conditioning regimes . There is no longer a need to specifically define the activation functions ν ( t ) , we instead rely solely on cross-correlation functions C ^ ( u ) and C ( u ) . Below , we aim to construct versions of these functions that are as close to experimental data as possible . Before discussing spiking statistics , we note an important advantage of only considering mean synaptic strengths J ¯ ( t ) . For spiking simulations shown above , we used networks of N = 60 neurons with probability of connection p = 0 . 3 , which are considerably far from realistic numbers . Nevertheless , the important quantity for mean synaptic dynamics is the averaged summed strengths of synaptic inputs that a neuron receives from any given group: p N 3 J ¯ α β . Notice that many choices of p and N can lead to the same quantity , therefore creating a scaling equivalence . Moreover , additional scaling of Jmax can further accommodate different network sizes . So far , we assumed that every neuron receives an average of 6 synapses from each group . If each of these synapse were at maximal value Jmax = 0 . 1 , then simultaneous spiking from a pre-synaptic group would increase the post-synaptic neuron’s spiking probability by 60% , a number we consider reasonable . It remains unclear if the mechanisms described above are consistent with the experimentally observed relationship between stimulation delay ( d† ) and efficacy of spike-triggered conditioning in macaque MC . We investigated this by comparing efficacy , as measured by the percentage of torque direction change evoked by ICMS before and after conditioning [10] , to relative synaptic strength changes in our model . This is motivated by the above demonstration that synaptic strengths are well correlated with amplitude of evoked muscle activations in a ICMS experiment ( see Fig 3 and point b . in Introduction ) . Nevertheless , the following comparison between model and experiment is qualitative , and meant to establish a correspondence of ( delay ) timescales only . We use the data originally presented in Figure 4 of [10] , describing the shift in mean change in evoked wrist torque direction by ICMS of the recorded site ( group a ) , as a function of stimulation delay d† . We plot the same data in Fig 4E , with the maximal change ( in degrees ) normalized to one . On the same graph , we plot the ( J ¯ b a † - J ¯ b a ) / J m a x v . s . d† curve for the value of σ that offers the best fit ( in L1-norm ) . This amounts to finding the best “σ-slice” of the graph in Fig 4D to fit the experimental data . We found that σ ≃ 17 ms gives the best correspondence . We reiterate that this comparison is qualitative . Nevertheless , the fit between the d†-dependence of experimentally observed functional changes and modelled synaptic ones is clear . As our model’s spiking activity and STDP rule are calibrated with experimentally observed parameters ( see Methods ) , this evidence suggests that our simplified framework is consistent with experiments . Importantly , σ = 17 ms is comparable to correlation timescales between functionally related MC neurons in macaque , as reported in [41] ( see also [5 , 40] ) and discussed earlier . It was shown that such neurons have gaussian-like correlation functions with mean peak width at half height on the order of 22 ms , corresponding roughly to σ = 10 ms . While this is slightly lower than our estimate , we note that task-specific motion is known to introduce sharp correlations and that free behaving , rest and sleep states induce longer-range statistics [42] . Cross-correlation functions reported above were recorded during a stereotyped center-out reach task experiment , in contrast to the spike-triggered conditioning experiment which was conducted over a wide range of states , including sleep , which may lead to longer mean cross-correlation timescales [10] . A prediction of our model is that spike-triggered conditioning restricted to periods of long timescale correlations in MC , such as during sleep [42] , could lead to a more robust conditioning dependence on stimulation delays ( see Discussion ) . This finding implies that our model successfully reproduces the third and last experimental observation from [10] ( point c . from Introduction ) : using simplified cross-correlation functions of MC neural populations calibrated from experimental measurements , our model reproduces the relationship between the magnitude of plastic changes and the stimulation delay in a spike-triggered conditioning protocol . We now explore the effects of spike-triggered stimulation on collateral synaptic strengths , i . e . , other than the targeted a-to-b pathway . For a wide range of parameters , there is little change other than for the a-to-b synapses . Indeed , when cross-correlation width σ is moderate to large , spike-triggered stimulation has little effect on collateral connections , for any stimulation delay . This is in conjunction with the robustness of a-to-b changes discussed in the previous section ( see Fig 4D ) . Nevertheless , some localized features arise when cross-correlation width σ is small . Fig 5A shows color plots of these changes as a function of d† and σ , for the nine combinations of pre- and post-synaptic groups . We now review the mechanisms responsible for these indirect synaptic alterations . First , b-to-b synapses become depressed , regardless of stimulation delay , for short correlation timescales . This is due to the occurrence of synchronized population spiking produced by artificial stimulation which , because of our choice of dendritic and axonal delays ( see Methods ) , promote depression . Such synchronized spikes induce sharp δ-peaks in the network’s cross-correlation ( see Methods ) and the combination of transmission delays shifts this peak toward the depression side of the STDP rule . When cross-correlations are narrow ( i . e . small σ ) , their interaction with the STDP rule –which manifests in the integral in Eq ( 13 ) – is more sensitive to the addition of such δ-peaks , resulting in overall depression . In contrast , when cross-correlations are wider , the addition of δ-peaks has a smaller effect since a wider range of correlations contribute to the integrated STDP changes . Second , the b-to-a synapses become potentiated for short delays d† when σ is small enough . This happens because of a combination of factors . When the recorded neuron in group a spikes , the population-wide spike artificially elicited in b quickly follows and travels to the b-to-a synapses . This means that the spike of a single neuron in a effectively reaches all neurons in a , with an effect amplified by the strength of many synapses , shortly after the neuron originally fired . When cross-correlations among a-neurons are wide , the effect of this mechanism is diluted , similarly to the b-to-b synapses discussed above . However , when neurons in a are highly synchronous , this short-latency feedback produces synaptic potentiation of the b-to-a synapses . Third , the synapses from both groups a and b onto the control group c are also potentiated when σ and d† are small enough . This can be explained in two parts and involves di- and tri-synaptic mechanisms . When the recorded neuron in a fires a spike , a population-wide spike is artificially evoked in b shortly after , which travels down to b-to-c synapses and elicits a response from neurons in c . Narrow cross-correlations imply that many spikes in a fall within a favorable potentiation window of spikes in c , thereby contributing to the potentiation of a-to-c synapses . To test this mechanism , we compute the relative synaptic changes due to spike-triggered stimulation in an altered network , where b-to-c synapses are inactivated ( Jcb ≡ 0 ) . Fig 5B shows the synaptic changes for the normal network ( top ) and this altered one ( middle ) for fixed parameters d† = 5 ms and σ = 17 ms ( same σ that best fitted experiments , see Fig 4E ) . We can clearly see that without b-to-c synapses , a-to-c synapses do not potentiate under spike-triggered stimulation . In turn , the strengthening of a-to-c synapses imply that spikes in a are more likely to directly elicit spikes in c , thereby repeating the same process in a different order for b-to-c synapses . Note that without a-to-c synapses , the b-to-c synapses would not potentiate . Indeed , as was the case for b-to-b synapses , the combination of transmission delays do not conspire to promote direct potentiation following a population-wide synchronous spike . This is tested by inactivating the a-to-c synapses , which prevents the potentiation of b-to-c synapses , as shown in the bottom panel of Fig 5B . Finally , there is a moderate increase of the c-to-b synapses for stimulation delays d† from about 5 to 20 ms . These are observed because of a-to-c synapses , promoting spikes in c that precede the stimulation of group b . This mechanism only works if neurons in a are tightly correlated , i . e . , for small σ . Using the same process described above , we tested this mechanism by inactivating a-to-c synapses which prevents the potentiation of c-to-b synapses . We reiterate that our specific choice of synaptic transmission delays may influence the magnitude of the changes described above . This is because many of the outlined mechanisms rely on well-timed series events . See e . g . [18 , 19 , 30 , 31] for more details about the impact of delays . We expect that tuning our model to experimentally measured delay distributions , as they become available , will help validate and/or improve our framework’s predictive power . Together , these mechanisms form our second novel finding ( point 2 . from the Introduction ) : multi-synaptic mechanisms give rise to significant changes in collateral synapses during conditioning with short delays , and these cannot be attributed to STDP mechanisms directly targeted by the BBCI , instead emerging from network activity . We note that the correlation time-scale that best fits experiments ( σ = 17ms , see Fig 4E ) falls within the parameter range where these effects occur . We have assumed that spike-triggered stimulation elicits population-wide synchronous spiking of all neurons in group b ( Nstim in [10] ) . This is valid if the neural group b represents all the neurons activated by the stimulating electrode of a BBCI , but is not necessarily representative of the larger population of neurons that share external activation statistics due to a common input νb ( t ) . Indeed , some neurons that activate in conjunction with those close to the electrode may be far enough from it so they do not necessarily spike in direct response to a stimulating pulse . Alternatively , selective activation of neurons within a group can also be achieved via optogenetic stimulation in a much more targeted fashion [43] . We now consider the situation in which only a certain proportion of neurons from group b is activated by the spike-triggered stimulus . We denote the stimulated subgroup by b† and the unstimulated subgroup by b∘ . All neurons in group b receive the same external rates νb ( t ) as before , but only a few ( solid red dots in Fig 6A ) are stimulated by the BBCI . Let Nb = N/3 be the number of neurons in group b and the parameter ρ , with 0 ≤ ρ ≤ 1 , represent the proportion of stimulated neurons in b . The sizes of groups b† and b∘ are given by N b † = ρ N b and N b ∘ = ( 1 - ρ ) N b , respectively . We now adapt our analytical averaged model ( 23 ) to explore the effect of stimulation on subdivided synaptic equilibria . We verified that the analytical derivations used below match the full spiking network simulations as before . Both subgroups of b receive the external rate νb ( t ) but only one receives spike-triggered stimulation . These changes are captured in the averaged analytical derivations by tracking the number of neurons in each sub-group in the averaging steps leading to Eqs ( 22 ) and ( 23 ) accordingly—replacing N/3 by N b † and N b ∘ where necessary . This way , we obtain subgroup-specific synaptic averages ( e . g . J ¯ b † a ) . Fig 6B shows the group-averaged connectivity strength between group a ( the recording site ) and both subgroups of b , before and after spike-triggered stimulation . External cross-correlations C ^ ( u ) are as in Fig 4B with σ = 20ms , and the stimulation delay is set at d† = 30ms . The proportion of stimulated neurons in b is set to ρ = 0 . 5 . The bottom of the same panel shows the normalized changes of mean synaptic strengths due to spike-triggered stimulation . As established for the original network ( see Fig 2B ) , the biggest change occurs for synapses from a to the subgroup that is directly stimulated ( b† ) . However for subgroup b∘ , we see a noticeable change in its incoming synapses from group a , in contrast to synapses of other unstimulated groups ( c ) that do not appreciably change . This means that sharing activation statistics with stimulated neurons is enough to transfer the plasticity-inducing effect of conditioning to a secondary neural population . Next , we investigate how this phenomenon is affected by the proportion of neurons in b that receive stimulation , ρ . Fig 6C shows the subgroup-averaged normalized changes of synapses from group a to subgroups b† and b∘ , as ρ varies between 0 and 1 . When more neurons get stimulated , the transferred effect on the unstimulated group is amplified . This means that the combined outcome on the entirety of group b grows even faster—supralinearly—with ρ , as shown in Fig 6C , where the combined b-averaged changes in synaptic strength , J ¯ b a = ρ J ¯ b † a + ( 1 - ρ ) J ¯ b ∘ a , are plotted as a function of ρ . In summary , this phenomenon represents our third and final finding ( point 3 . in Introduction ) . Our model shows that neurons not directly stimulated during spike-triggered conditioning can be entrained into artificially induced plasticity changes by a subgroup of stimulated cells , and that the combined population-averaged effect grows supra-linearly with the size of the stimulated subgroup .
In this study , we used a probabilistic model of spiking neurons with plastic synapses obeying a simple STDP rule to investigate the effect of a BBCI on the connectivity of recurrent cortical-like networks . Here the BBCI records from a single neuron within a population and delivers spike-triggered stimuli to a different population after a set delay . We developed a reduced dynamical system for the average synaptic strengths between neural populations; these dynamics admit stable fixed points corresponding to synaptic equilibria that depend solely on the network activity’s correlations and spike-triggered stimulation parameters ( see Methods ) . In this framework , individual synapses may fluctuate with ongoing network activity but their population average remains stable in time . We validate our findings with detailed numerical simulations of a spiking network and calibrate our result based on experiments in macaque MC . To our knowledge , this is the first time a plastic spiking network model includes recurrent network interactions to capture the effects of a BBCI . We successfully reproduce key experimental results from [10] that describe synaptic changes due to spike-triggered conditioning . Specifically , we recover the two emergent timescales with which these changes occur ( hours to days ) , we show that filtered evoked activity from our network mimics muscle EMG patterns produced by ICMS protocols , and we show that maximal changes in mean synaptic strength from the recording site to the stimulated site occur with stimulation delays in the 20–50 ms range , as was the case for experiments . Furthermore , we formulate three novel findings using our model . We outline the relationship between temporal statistics within neural populations and optimal stimulation parameters , we uncover multi-synaptic mechanisms that emerge from spike-triggered conditioning that are not directly predicted by a single-synapse STDP rationale , and we find that the stimulation of a subset of neurons within a population can lead to supra-linear scaling of effects . Based on this , we formulate two main experimental predictions: In our model , common external inputs are needed to endow neural groups with desired statistics . In cortical networks , it is unclear how much of observed activity is driven from external sources and how much is due to intra-network interactions . We argue that the use of prescribed external activations is appropriate since we show that plasticity works to “learn” activation correlations ( see also [18–22] ) , thereby guiding synaptic connections to promote the same spontaneous network statistics as its driven ones . Therefore , while external input rates are necessary in our model , the resulting network correlations C ( u ) reflect both emergent and external dynamics , and serve as a proxy for MC activity . An interesting outcome of our findings is that spike-triggered conditioning is strongly influenced by the source and target cross- and auto-correlation structure C ( u ) . Throughout the paper , we assumed that these statistics were stationary for the entire simulated period . However , it is well known that cortical circuits can show a wide range of activity regimes depending on the state of the animal . The statistics of MC neurons may be very different if the animal is awake and behaving freely or performing a precise motor task , or is asleep [42] . While such states have limited durations , their timescales ( from minutes to hours ) may be long enough to define locally-stationary statistics that BBCI protocols could leverage to optimize desired effects . For example , stimulation during sleep , which is known to produce oscillation-rich activity with longer-range correlations [42 , 44] , could have more robust but slower effects , while stimulation during a specific task can have more targeted outcomes . Our model can easily be scaled up to include multiple recording and stimulation sites , and different stimulation protocols , including , e . g . EMG-triggering [45] or paired-pulse stimulation [46] . It is also easily adaptable to optically-based stimulation which can target specific neurons within functional groups ( see e . g . [43 , 47 , 48] ) . As it does not require costly simulations—only theoretical estimates—it is straightforward to apply optimization algorithms to find the best stimulation protocol to achieve a desired connectivity between cortical sites . Furthermore , it can easily incorporate closed-loop signals such as changes in recorded statistics in real-time . This framework offers a flexible testbed to help design experiments and clinical treatments . Neural implants such as BCIs and BBCIs are under active development as they have significant potential for clinical use [49] . Among other outcomes , they are promising avenues for treatment of motor disabilities . Indeed , a BBCI capable of inducing plastic changes in cortical circuits could be used to promote novel synaptic pathways in order to restore functional connectivity after a stroke or injury [50 , 51] . For such bidirectional neural implants to be successful , a number of real-time computational issues need to be resolved . Our modelling framework presents a step toward the development of rapidly computable guides for controllers , based on anatomical organization , measurable network properties and known physiological mechanisms such as STDP . While being able to capture important features , our model misses some important physiological aspects of MC circuits . Nevertheless , we believe that the simplicity of our model is a strength: it captures complex network-level plasticity changes with only excitatory activity and STDP . This suggests that excitatory mechanisms are central to artificially-induced plasticity by a BBCI . However , this simplicity will likely be insufficient to reproduce more complex protocols that also recruit inhibitory populations . An example is paired-pulse conditioning , where the BBCI stimulates several neural sites , with different time delays . Additional biological realism is key to expand our bottom-up theoretical framework . A number of steps can be taken to add physiological realism , each of them adding some complications . For inhibition , there are technical issues when considering probabilistic spiking ( e . g . inhibition can induce “negative” spiking probabilities if unchecked ) , and the STDP mechanisms for inhibitory synapses are not well understood , although this is changing rapidly ( see e . g . [52 , 53] ) . The implementation of our framework in a dynamical model setting , building on theoretical models of inhibitory plasticity ( following e . g . [54 , 55] ) are natural next steps . However , with added realism comes additional complexity , and it is not clear if analytically tractable results can be derived . Nevertheless , preliminary numerical simulation of our model , with added inhibition shows that although novel mechanisms emerge from inhibitory dynamics , the qualitative phenomena described in this article persist . In addition , the inclusion of spatial structure and heterogeneous delay distributions , based on anatomical data [9] , is necessary to expand the framework to multiple cortical sites and spinal cord . There is also evidence that more complex excitatory synaptic rules involving spike timing , such as the “triplet rule” [56] and history dependent rules [57] , may play important roles in cortex . Such synaptic mechanisms involve more complex statistical interactions between neurons and even between a neuron and its spiking history . Mathematical methods that extract the higher order statistics needed to derive the self-consistency equations for synaptic equilibria are currently being developed ( e . g . [58] ) , and their application to BBCI modelling frameworks similar to ours is a promising avenue for future work . Finally , the inclusion of modulatory mechanisms , activity- or chemically-dependent , is crucial to capture phenomena such as synaptic consolidation and adaptation . In summary , our reduced dynamical system approach is a promising basis upon which to build and ultimately to predict the effects of finer-grained cell-type specific and temporally structured activation patterns afforded by next-generation neural implants using electrical or optical stimulation .
Our goal is to derive an analytical expression for the dynamics of the synaptic matrix J ( t ) , to allow us to predict the timescale of plastic changes and relative equilibria . Eqs ( 2 ) and ( 6 ) show that changes of the synapses stored in J ( t ) , modulated by the spiking activity of the network , depend on the external rates ν ( t ) and the synaptic matrix J ( t ) itself . This leads to a self-consistent relationship between J ( t ) dynamics and network spiking activity . Earlier work [18–22] describes a framework to effectively decouple these interactions , which we follow and adapt to our needs . The key idea , originally proposed in [16] , is to use a separation of timescales , assuming that synapses are locally constant over time intervals [t1 , t2] , on which the network has stable and stationary spiking statistics . It follows that accumulated increments for the synapse Jij over that interval , denoted ΔJij = Jij ( t2 ) − Jij ( t1 ) , is given by the sum of all discrete plastic “steps” due to spike-time pairs t i s , t j s ∈ [ t 1 , t 2 ] arriving at the ij-synapse ( ignoring delays for clarity ) : Δ J i j = η ∑ t 1 ≤ t i s , t j s ≤ t 2 W ( t j s - t i s , J i j ( t 1 ) ) . ( 8 ) An equivalent formulation of Eq ( 8 ) uses the density of interspike intervals u = t j s - t i s over the interval [t1 , t2] . To this end , let Cij ( u ) be the count of spike pairs separated by u occurring over the interval [t1 , t2] , which is the ( non-normalized ) cross-correlation between neuron j and neuron i . Then , Δ J i j = η ∫ - ∞ ∞ d u C i j ( u ) W ( u , J i j ( t 1 ) ) . ( 9 ) Eq ( 9 ) outlines the basic mechanism governing the evolution of the synaptic matrix J ( t ) over consecutive time intervals . The challenge is to express Cij ( u ) —itself nontrivially dependent on J ( t ) and external rates ν ( t ) —in a closed form , so that Eq ( 9 ) can be iterated by updating J at each step . The authors of [18] present detailed descriptions of this iteration process , and of ways to estimate the cross-correlations Cij ( u ) . However , a number of limitations of this original derivation prevent us from using it directly: ( i ) The assumption that external rates νi ( t ) and νj ( t ) are δ-correlated; we require arbitrary cross-correlation functions to fit the model to experiments . ( ii ) Artificial spike-triggered stimulation induced by the BBCI introduces non-trivial statistical dependencies between groups and single neurons . In the following , still closely following [18] , we present a theoretical derivation that addresses these issues . First , we describe parametric constraints that ensure that spiking activity remains stable and does not run away in a self-excitation cascade . Second , we formally describe the timescale separation argument outlined above , followed by estimates for network cross-correlations , both for normal activity and in the presence of spike-triggered stimulation . Finally , we formulate and analyze a dynamical system for averaged synaptic dynamics . In Results , we use experimentally obtained cross-correlation functions from spiking activity recorded in a monkey implanted with a Utah array in primary MC and performing an isometric 2D target-tracking task . Spikes from multiple electrodes of the array were recorded during task performance , sorted , and cross-correllograms were compiled with a resolution of 2 milliseconds . Recorded activity was modulated during task performance . We assume the neurons whose spikes are recorded from the same electrode ( using spike sorting ) were close enough to be part of the same neural group , in the context of our model . In contrast , neurons recorded from different electrodes , separated by at least 400 μm , are considered as part of different groups . | Recent developments in Bidirectional Brain-Computer Interfaces ( BBCI ) not only allow the reading out of neural activity from cortical neurons , but also the delivery of electrical signals . These drive neural dynamics and shape synaptic plasticity , thus opening the possibility of engineering novel neural circuits , with important applications for clinical treatments of spinal cord injuries and stroke . However , synaptic changes in recurrent networks of neurons are hard to predict: they involve complex dynamic mechanisms on multiple temporal and spatial scales . Based on experiments , we develop a computational network model with plastic synapses that serves as a predictive tool for BBCI protocols . We show how the efficacy of BBCIs is influenced by cortical activity statistics and we propose state-based stimulation strategies for driving artificially-induced synaptic plasticity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"neural",
"networks",
"nervous",
"system",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"synaptic",
"plasticity",
"bioassays",
"and",
"physiological",
"analysis",
"muscle",
"electrophysiology"... | 2017 | Correlation-based model of artificially induced plasticity in motor cortex by a bidirectional brain-computer interface |
Candida albicans yeast cells are found in the intestine of most humans , yet this opportunist can invade host tissues and cause life-threatening infections in susceptible individuals . To better understand the host factors that underlie susceptibility to candidiasis , we developed a new model to study antifungal innate immunity . We demonstrate that the yeast form of C . albicans establishes an intestinal infection in Caenorhabditis elegans , whereas heat-killed yeast are avirulent . Genome-wide , transcription-profiling analysis of C . elegans infected with C . albicans yeast showed that exposure to C . albicans stimulated a rapid host response involving 313 genes ( 124 upregulated and 189 downregulated , ∼1 . 6% of the genome ) many of which encode antimicrobial , secreted or detoxification proteins . Interestingly , the host genes affected by C . albicans exposure overlapped only to a small extent with the distinct transcriptional responses to the pathogenic bacteria Pseudomonas aeruginosa or Staphylococcus aureus , indicating that there is a high degree of immune specificity toward different bacterial species and C . albicans . Furthermore , genes induced by P . aeruginosa and S . aureus were strongly over-represented among the genes downregulated during C . albicans infection , suggesting that in response to fungal pathogens , nematodes selectively repress the transcription of antibacterial immune effectors . A similar phenomenon is well known in the plant immune response , but has not been described previously in metazoans . Finally , 56% of the genes induced by live C . albicans were also upregulated by heat-killed yeast . These data suggest that a large part of the transcriptional response to C . albicans is mediated through “pattern recognition , ” an ancient immune surveillance mechanism able to detect conserved microbial molecules ( so-called pathogen-associated molecular patterns or PAMPs ) . This study provides new information on the evolution and regulation of the innate immune response to divergent pathogens and demonstrates that nematodes selectively mount specific antifungal defenses at the expense of antibacterial responses .
Candida albicans is a remarkably successful and versatile human pathogen that is found on the skin and mucosal surfaces of virtually all humans . Under most circumstances , C . albicans is a harmless commensal [1] . However , this opportunist can invade host tissues and cause life-threatening infections when the immune system is weakened ( e . g . from critical illness ) and competing bacterial flora are eliminated ( e . g . from broad-spectrum antibiotic use ) . Accordingly , invasive candidiasis is particularly common in intensive care units where mortality rates reach 45–49% [2]–[4] . Antecedent colonization of mucosal surfaces with C . albicans can also lead to debilitating superficial infections in otherwise normal hosts . Approximately 75% of all women , for example , will have one episode of Candida vaginitis in their lifetime , with half having at least one recurrence [5] . C . albicans can grow vegetatively as yeast or hyphae , and each form contributes to pathogenesis [6]–[8] . C . albicans yeast cells colonize mucosal surfaces and facilitate dissemination of the organism through the blood stream [9]–[11] . Hyphae , by contrast , are important for host invasion and tissue destruction [1] , [8] , [11] , [12] . The factors that influence these diverse growth patterns during infection are poorly understood , but it is clear that innate immune mechanisms in mammalian epithelial cells normally prevent C . albicans from becoming a pathogen [13]–[15] . Recently , genetic analyses of two human families whose members suffered from recurrent or chronic candidiasis on mucosal surfaces identified causative mutations in the innate immune regulators dectin-1 [16] and CARD9 [17] . Dectin-1 is a pattern-recognition receptor important for macrophage phagocytosis of fungi . Interestingly , this protein interacts differently with the C . albicans growth forms . Cell wall components exposed in the bud scar of C . albicans yeast ( so-called pathogen-associated molecular patterns or PAMPs ) potently stimulate dectin-1 , but hyphae are relatively shielded from innate immune detection , which likely contributes to the ability of C . albicans to establish infection [13] , [15] , [18] . Furthermore , a recent study found that the p38 MAP kinase , a central regulator of mammalian immunity , receives biphasic inputs from C . albicans that are dependent on the morphologic form of the organism and the local fungal burden [14] . These data suggest that the interplay between C . albicans and the mammalian innate immune system dictate the virulence potential of this specialized pathogen , yet relatively little is known about the molecular mechanisms underlying these interactions . One approach to study evolutionarily conserved aspects of epithelial innate immunity and microbial virulence uses the invertebrate host Caenorhabditis elegans [19] , [20] . In nature , nematodes encounter numerous threats from ingested pathogens , which have provided a strong selection pressure to evolve and maintain a sophisticated innate immune system in its intestinal epithelium [21] . Coordination of these defenses involves several highly-conserved elements that have mammalian orthologs [22]–[25] . Furthermore , C . elegans intestinal epithelial cells bear a striking resemblance to human intestinal cells [26] and because the nematode lacks both a circulatory system and cells dedicated to the immune response , the intestinal epithelium constitutes the primary line of defense for the nematode against ingested pathogens . Thus , it is possible to conduct analyses of innate immune mechanisms in a physiologically-relevant , genetically-tractable system . Much of the characterization of nematode immunity has used nosocomial bacterial pathogens [27]–[30] , particularly Pseudomonas aeruginosa [22] , [31] , [32] , but to date , the immune response directed toward a medically-important , fungal pathogen has not been defined . Here , we extend our previously-validated system for the study of hyphal-mediated C . albicans virulence in the nematode [33] to examine C . albicans yeast . Our goal was to use studies of C . elegans-C . albicans interactions to identify novel , conserved features of metazoan innate immunity . We found that the responses to bacterial and fungal pathogens are remarkably distinct . Many of the immune response effectors that are upregulated by either P . aeruginosa or S . aureus are downregulated by infection with C . albicans yeast . We also found that slightly more than half of the immune response genes activated by infection with live C . albicans are also upregulated by heat-killed C . albicans . Our data indicate that the C . elegans immune response to C . albicans most likely involves detection of conserved surface-associated molecular pattern molecules , as well as detection of C . albicans virulence-related factors .
To examine interactions between C . albicans and the innate immune system , we established a novel system using the model host C . elegans . In a previous study , we found that C . albicans hyphae can kill C . elegans in a manner that models key aspects of mammalian pathogenesis [20] , [33] . In that assay , yeast cells were ingested by nematodes on solid medium and , after transfer to liquid medium , worms died with true hyphae piercing through their bodies . During these experiments , we noted that when infected worms were maintained on solid media , rather than transferred to liquid media , the C . albicans yeast form caused pathogenic distention of the nematode intestine and premature death of the worms . Thus , we hypothesized that C . albicans yeast , the form commonly found in the mammalian intestine [13] , [15] , [18] , also contain virulence determinants that allow infection of C . elegans . We therefore developed an assay that is conducted exclusively on solid media and allows the direct study of yeast-mediated pathogenesis of the nematode . As shown in Figure 1 , the yeast form of the C . albicans laboratory reference strain DAY185 infected and killed C . elegans . Heat-killed C . albicans yeast cells were not pathogenic to the nematode ( Figure 1A ) and caused less distention of the nematode intestine compared to that seen following exposure to live C . albicans ( Figure 1B ) . We found that the C . albicans clinical isolate SC5314 was also able to establish a lethal infection in nematodes ( Figure 2 ) . Furthermore , the C . albicans efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant strain [8] , which is attenuated for virulence in mammals , was also unable to efficiently kill C . elegans in this assay ( Figure 2 ) . Like its isogenic wild-type parent strain , virulence-attenuated C . albicans yeast enter the nematode intestine during the infection assay ( data not shown ) , suggesting that non-specific occlusion of the intestine with yeast is not the mechanism of C . albicans-mediated worm killing . In addition , we found that C . albicans killed sterile C . elegans fer-15 ( b26 ) ;fem-1 ( hc17 ) animals ( data not shown ) and wild-type worms in the presence of 5-fluoro-2′-deoxyuridine ( FUDR ) , a compound that prevents progeny from hatching ( Figure 1A ) . These results suggest that killing of nematodes by C . albicans yeast in the C . elegans model involves virulence determinants intrinsic to live fungi and not a “matricidal effect” from premature hatching of embryos inside animals , a previously described , non-specific consequence of pathogen stress in wild-type worms [26] , [31] , [32] , [34] . In summary , these data demonstrate that C . albicans yeast are pathogenic to the nematode and establish a second assay , which together with the liquid-media system [33] , permit separate in vivo analyses of C . albicans growth states . Previous studies have shown that C . elegans mounts a rapid and specific immune response toward pathogenic bacteria [32] , [35] , [36]; however , it is not known how the nematode defends itself against an intestinal fungal pathogen . We therefore used transcriptome profiles of nematodes during an infection with C . albicans yeast to define the antifungal immune response genes in the nematode . We compared gene expression of animals exposed to C . albicans for four hours with control worms fed the non-pathogenic food source , heat-killed E . coli OP50 . The short exposure time maximized the yield for transcriptional changes associated with pathogen detection , rather than gene expression changes associated with intestinal damage [36] . It was necessary to use heat-killed E . coli for these experiments because live E . coli were previously shown to be pathogenic to the nematode on C . albicans growth media ( brain heart infusion agar ) [37] . We found that C . elegans coordinates a rapid and robust transcriptional response to C . albicans that involves approximately 1 . 6% of the nematode genome ( Figure 3 ) . 124 genes were upregulated two-fold or greater in response to C . albicans compared to heat-killed E . coli and 189 genes were downregulated at least two-fold ( P<0 . 01 ) ( Figure 3A and Table S1A ) . For technical confirmation of the microarray experiment , we selected 11 genes that showed varying degrees of differential regulation and tested their expression by quantitative real-time polymerase chain reaction ( qRT-PCR ) under each microarray condition ( Figure 3B and Table S2 ) . Plotting the fold difference observed in the transcriptome profiles versus the value obtained by qRT-PCR from the three biological replicates used for the microarray analysis yielded an R2 of 0 . 90 ( Figure 3B ) , which indicates tight correlation between these datasets and is a result that compares favorably with similar analyses of other microarray experiments [38] . We also tested three additional biological replicates and found similar fold changes between the microarray and qRT-PCR analyses in 10 of the 11 genes ( Table S2 ) , a correlation rate that is consistent with other microarray analyses of pathogen response genes in the nematode [34] . As a third means to confirm the results of our microarray , we compared the expression of 4 upregulated and 4 downregulated genes in wild-type C . elegans animals infected with a different C . albicans strain than used for the microarray analysis . We exposed animals to the C . albicans clinical isolate SC5314 , a strain that is also virulent toward C . elegans ( Figure 2 ) , and found similar transcriptional changes between C . albicans SC5314 and DAY185-exposed animals for all 8 genes tested ( Table S2 ) . These data suggest that the C . albicans-induced transcriptional changes observed in our microarray analysis are not specific to a particular yeast strain . Examination of the genes induced by C . albicans in the microarray analysis reveals the footprint of an immune response toward a pathogenic fungus ( Table 1 ) . C . albicans infection results in the elaboration of at least seven putative antimicrobial peptides , which are postulated to have antifungal activity in vivo . One of these genes , abf-2 , was previously shown to have in vitro activity against the pathogenic fungus Candida krusei [39] . Three genes in this group ( fipr-22/23 and two caenacin genes , cnc-4 and cnc-7 ) are antifungal immune effectors induced by the nematode following exposure to Drechmeria coniospora , an environmental fungal pathogen , which causes a localized infection of the nematode cuticle [40] , [41] . fipr-22 and fipr-23 have nearly identical DNA sequences and thus , it is not possible for a probe set to distinguish between these genes . Two chitinase genes ( cht-1 and T19H5 . 1 ) were also strongly induced by C . albicans . These enzymes are secreted by metazoans and are thought to defend against chitin-containing microorganisms such as C . albicans and other pathogenic fungi [42] , [43] . In addition , thn-1 , a gene that is postulated to have direct antimicrobial activity and is a homolog of the thaumatin family of plant antifungals [35] , [44] , was induced 2 . 5-fold during infection with C . albicans . Using gene expression analyses , we characterized further the expression pattern of four putative antifungal immune effectors upregulated during C . albicans infection ( abf-2 , fipr-22/23 , cnc-4 and cnc-7 ) . We exposed wild-type nematodes to the C . albicans efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant , a strain that is attenuated for virulence in C . elegans ( Figure 2 ) and mammals [8] , and found that the induction of abf-2 , fipr-22/23 , cnc-4 and cnc-7 was reduced compared to its isogenic parent strain C . albicans SC5314 ( P<0 . 01 for fipr-22/23 and cnc-7 , P = 0 . 06 for abf-2 , P<0 . 025 for cnc-4 ) ( Figure 4 ) . These data suggest that the nematode modulates the expression levels of antifungal immune effectors in response to some aspect of C . albicans virulence , although this yeast may be recognized differently by the nematode innate immune system owing to pleotropic effects of the genetic lesions in this mutant strain . We also found that the induction levels of these four genes appear to be dynamic during infection . Twelve hours after exposure to C . albicans , the expression of abf-2 increases significantly , fipr-22/23 is unchanged and cnc-4 and cnc-7 is reduced ( Figure S1 ) . Among the most highly upregulated C . albicans defense genes ( Table 1 ) , we also identified a preponderance of genes encoding secreted proteins , intestinally-expressed proteins and proteins that may function as detoxifying enzymes . Similar types of genes are induced following infection with pathogenic bacteria [32] , [34] . As discussed in more detail below , we also found that some of the C . albicans-induced genes were involved in the nematode transcriptional response to bacterial pathogens ( Table 1 ) , suggesting that C . albicans and pathogenic bacteria induce a set of common immune response effectors . Although it is possible that the effects of nematode starvation are also reflected in the transcription profiling data as a potential consequence of C . albicans being comparatively non-nutritious relative to heat-killed E . coli , this seems less likely since zero of the eighteen previously-identified , fasting-affected genes [45] were differentially expressed in the dataset . Taken together , these data suggest that the microarray analysis captured the early defense response mounted by C . elegans toward an ingested fungal pathogen . Genetic , biochemical and molecular analyses have identified a requirement for the PMK-1 mitogen-activated protein ( MAP ) kinase , orthologous to the mammalian p38 MAPK , in C . elegans immunity [22] , [29] , [46]–[48] . PMK-1 is a central regulator of nematode defenses [32] that acts cell autonomously both in the intestine to control resistance toward the Gram-negative bacterial pathogens P . aeruginosa [47] and Yersinia pestis [29] , and in the hypodermis to defend against the fungus D . coniospora [46] . We found that C . elegans pmk-1 ( km25 ) mutants were hypersusceptible to infection with C . albicans yeast ( Figure 5A ) and that PMK-1 was required for the basal and pathogen-induced expression of three antifungal immune effectors ( fipr-22/23 , cnc-4 and cnc-7 ) , but not abf-2 ( Figure 5B ) . The full spectrum of nematode sensitivity to C . albicans was not mediated by the genetic control of any of these four effectors because knockdown of each of these genes individually by RNA interference did not result in hypersusceptibility to fungal infection ( data not shown ) . It is likely , however , that there is functional redundancy among immune effectors in C . elegans , as has been suggested previously [29] , [32] , [44] , [49] , [50] . That PMK-1 mediates resistance to C . albicans provides another line of evidence that yeast infection of the nematode stimulates host immune defenses . Moreover , the PMK-1-independent genetic regulation of the antifungal effector abf-2 suggests that other pathways are also important in controlling the immune response toward C . albicans . To examine the specificity of the antifungal transcriptional response , we compared C . albicans-affected genes with those differentially regulated following infection with the bacterial pathogens P . aeruginosa [32] and Staphylococcus aureus [34] ( P<0 . 01 , >2-fold change ) ( Figure 6 ) . The transcriptional responses induced by fungi , Gram-negative bacteria and Gram-positive bacteria overlapped only to a small extent and the majority of the C . albicans-affected genes were not involved in the response to P . aeruginosa or S . aureus ( Figure 6 , Table S3A ) . The C . albicans-specific genes in this comparison included the putative antifungal peptides abf-2 , fipr-22/23 , cnc-7 , thn-1 and the chitinases ( cht-1 and T19H5 . 1 ) . We observed an overlap of 32 induced and 22 repressed genes between the transcriptional responses to P . aeruginosa and C . albicans ( 1 . 9 and 1 . 4 genes expected by chance alone , respectively; P<1 . 0×10−16 for both comparisons ) . Likewise , 22 upregulated and 25 downregulated genes were shared in the responses to S . aureus and C . albicans ( 2 . 8 and 2 . 2 genes expected by chance alone , respectively; P<1 . 0×10−16 for both comparisons ) . Interestingly , 12 genes were induced and 14 genes were repressed by all three pathogens . Despite the fact that the C . albicans-induced genes were determined using heat-killed E . coli as the control and the genes induced by P . aeruginosa and S . aureus were identified in separate studies that used live E . coli as the control , we detected an overlap of comparable significance between the transcriptional responses to these different organisms . 26% and 18% of C . albicans-induced genes were also upregulated by P . aeruginosa and S . aureus , respectively ( Figure 6 ) . Likewise , 17% of genes induced by P . aeruginosa four hours after infection were also upregulated by S . aureus and 11% of S . aureus-upregulated genes were induced by M . nematophilum [34] . Our data suggest that the nematode is able to specifically recognize C . albicans infection and mount a targeted response toward this fungus that involves antifungal defenses and a limited number of common core effectors . Components of the C . albicans cell wall , often referred to as PAMPs , are recognized by mammalian neutrophils , monocytes and macrophages [13] , [15] , [51] . In this study , we found that heat-killed C . albicans yeast accumulate within the C . elegans intestine ( Figure 1B ) and therefore postulated that the nematode transcriptional response to nonpathogenic , heat-killed fungi would reflect stimulation of host pathways by immunogenic components of the yeast cell wall . To explore the mechanisms of pathogen detection in the nematode , we fed animals heat-killed C . albicans as an additional condition in the transcriptome profiling experiment . Exposure to heat-killed C . albicans caused a transcriptional response in nematodes involving 287 genes ( ∼1 . 4% of the genome , P<0 . 01 ) ( Table S1B ) . To determine whether these genes were also involved in defense against live C . albicans infection , we compared the genes differentially regulated by live and heat-killed C . albicans versus the baseline condition of heat-killed E . coli . Interestingly , there was significant overlap ( 69 genes , 56% ) between genes induced by heat-killed C . albicans ( vs . heat-killed E . coli ) and live C . albicans ( vs . heat-killed E . coli ) ( 0 . 5 genes expected by chance alone , P<1 . 0×10−16 ) ( Figure 7A , Table S3B ) . Likewise 106 of 189 genes ( 56% ) repressed by C . albicans were also downregulated by heat-killed C . albicans ( 0 . 5 genes expected by chance alone , P<1 . 0×10−16 ) ( Figure 7B , Table S3B ) . Interestingly , this overlap includes the majority of the most strongly regulated genes in both directions ( Tables 1 and S1A ) . These data constitute the first genome-wide analysis of the C . elegans transcriptional response to a heat-killed pathogen and afford several interesting observations . Heat-killed C . albicans yeast cells induce an antifungal transcriptional response in C . elegans despite being non-pathogenic ( Figure 1 ) . Genes upregulated by heat-killed C . albicans include several putative antifungal peptides ( abf-2 , cnc-4 , cnc-7 , cht-1 and thn-1 ) and an abundance of secreted or intestinal expressed genes ( Table 1 ) , a profile similar to that of live C . albicans . Furthermore , heat-killed C . albicans caused the induction of core immune response genes . The comparison in Figure 6 showed that 42 genes were upregulated by C . albicans and either P . aeruginosa or S . aureus . Thirty-three genes ( 79% ) in this set , including 7 out of 12 genes induced by all three pathogens , were also upregulated by heat-killed C . albicans ( Table S3A ) . Together , these findings suggest that heat-killed C . albicans yeast induce host defenses and imply that a large part of the C . elegans transcriptional response may be mediated by detection of fungal PAMPs through Pattern Recognition Receptors , an evolutionarily-ancient system of pathogen sensing and signaling [52] , [53] . Equally interesting , it seems that C . elegans also possesses mechanisms to respond directly to the virulence effects of C . albicans . We identified a smaller group of differentially regulated genes when we compared the transcriptome profiles from nematodes exposed to live C . albicans with those exposed to heat-killed C . albicans . The transcription of 62 genes ( 22 upregulated and 40 downregulated ) changed in this analysis ( P<0 . 01 ) ( Table S1C ) presumably in response to the pathogenicity of the fungus . 10 of the 22 genes ( 45% ) upregulated by live C . albicans versus heat-killed C . albicans and 11 of the 40 downregulated genes ( 28% ) were also differentially regulated by live C . albicans versus the baseline condition of heat-killed E . coli ( 0 . 12 and 0 . 36 genes respectively expected by chance alone , P<1 . 0×10−16 for both comparisons ) . These data are consistent with our observation that the induction of four putative antifungal effectors was reduced in the virulence-attenuated C . albicans efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant strain compared to its isogenic , wild-type parent strain ( Figure 4 ) . Taken together , these data indicate that host recognition of C . albicans infection in the nematode involves at least two mechanisms: recognition of PAMPs and detection of factors associated with fungal virulence . Closer examination of the genes downregulated by C . albicans revealed an unexpected finding regarding antifungal immune specificity . We noticed that the most over-represented classes among the C . albicans downregulated genes ( based on GO annotation ) were involved in sugar or carbohydrate binding . Because these gene classes are upregulated in response to P . aeruginosa and S . aureus [32] , [34] , we postulated that some antibacterial defense effectors are specifically downregulated during infection with C . albicans . We therefore compared the 189 genes that are downregulated by C . albicans with the genes induced during infection with P . aeruginosa and S . aureus , and found a striking overlap ( Figure 8A , Table S3C ) . Twenty-seven of the 189 downregulated C . albicans genes ( 14% ) were induced by P . aeruginosa , which is 25-fold more than expected by chance alone ( P<1 . 0×10−16 ) . Likewise , 22 S . aureus response genes ( 12% ) were downregulated by C . albicans ( 12-fold more than expected by chance alone , P<1 . 0×10−16 ) . Thus , it seems that the nematode immune response to C . albicans involves the downregulation of a group of antibacterial defense genes . We took two steps to confirm this observation . First , we used qRT-PCR to test the expression of seven genes differentially regulated by C . albicans and previously shown to be part of the P . aeruginosa transcriptional response ( irg-3 , clec-67 , K08D8 . 5 , C17H12 . 8 , F49F1 . 6 , F35E12 . 5 and F01D5 . 5 ) [32] . All seven of these genes were strongly downregulated four hours after C . albicans infection ( Table S2 ) . We also assayed the expression of clec-67 , K08D8 . 5 , C17H12 . 8 and F49F1 . 6 12 hours after infection and found that these genes continue to be transcriptionally repressed at this later time point ( Figure S1 ) . Two of these genes , C17H12 . 8 and F49F1 . 6 , were more strongly repressed at 12 hours compared to 4 hours after infection ( P<0 . 01 and P = 0 . 07 , respectively ) . As a second approach , we studied transgenic C . elegans animals in which the promoter for the S . aureus immune response gene clec-60 was fused to GFP , allowing a visual readout of gene transcription . clec-60 is a C-type lectin , a gene class important for nematode defense against bacterial pathogens [29] , [32] , [34] , a member of which was shown to have direct antimicrobial activity in a mammalian system [54] . Consistent with the microarray analysis ( Table S1A ) , we found that exposure to live C . albicans dramatically reduced GFP expression in clec-60::GFP transgenic animals compared to the basal expression of this gene on heat-killed E . coli ( Figure 8B ) . One interpretation of these data is that the downregulation of antibacterial effectors observed in the microarray analysis reflects the absence of bacteria in C . albicans-exposed animals rather than specific transcriptional repression of these genes during infection with pathogenic fungi . We therefore examined the genes that were downregulated in the comparison of live C . albicans versus heat-killed C . albicans , an experiment where bacterial antigens were not present in either condition . Of the 40 genes that were transcriptionally repressed in this comparison , 19 genes were also upregulated by S . aureus [34] or P . aeruginosa [32] ( Table S1C ) ( 0 . 08 genes expected by chance alone , P<1 . 0×10−6 for this comparison ) . For reasons that are not clear , only 6 of these 19 genes were also downregulated in the comparison of live C . albicans versus heat-killed E . coli ( Table S3C ) ; however , this overlap is significantly more than the 0 . 08 gene overlap expected by chance alone ( P = 0 . 013 ) . Therefore , we conclude that the nematode downregulates a group of antibacterial defense genes in response to some aspect of C . albicans virulence . It is also interesting that of the 44 antibacterial response genes shown in Figure 8 that were downregulated by C . albicans , 26 ( 59% ) were also repressed by heat-killed C . albicans ( Table S3C ) . Taken together , these data suggest that the nematode responds to components within heat-killed C . albicans , as well as factors associated with fungal virulence , to transcriptionally repress antibacterial immune responses . One of the antibacterial genes downregulated in the comparison of live C . albicans and heat-killed C . albicans was clec-60 . Thus , for additional confirmation of these data , we exposed clec-60::GFP transgenic animals to heat-killed C . albicans . As predicted from the microarray analysis , we found that expression of clec-60::GFP was visually unchanged compared to its basal level on heat-killed E . coli ( Figure 8B ) . Furthermore , our finding that C17H12 . 8 and F49F1 . 6 were more strongly downregulated at 12 hours of infection ( versus 4 hours ) ( Figure S1 ) suggests that the transcriptional repression of these antibacterial immune effectors is an active process associated with progression of fungal infection . To understand the mechanism underlying the repression of antibacterial immune effectors during C . albicans infection , we assayed gene expression in daf-16 ( mgDf47 ) and pmk-1 ( km25 ) mutants . Troemel et al . previously showed that the p38 MAP kinase homolog PMK-1 controls the expression of many P . aeruginosa immune response genes [32] . In their analysis , they also observed that the FOXO/forkhead transcription factor DAF-16 , a central regulator of nematode longevity , negatively regulates some P . aeruginosa defense genes , including a group of pmk-1-dependent genes . We therefore wondered whether DAF-16 negatively regulates antibacterial defense genes during infection with C . albicans . We determined the overlap of the C . albicans downregulated genes with the group of genes whose basal expression is negatively regulated by DAF-16 ( so-called Class II genes from Murphy et al . [55] ) and found a 24-gene overlap ( more than the 2 . 6 genes expected by chance alone , P<1 . 0×10−16 ) . From these analyses , we identified two genes ( clec-67 and C17H12 . 8 ) whose basal expression was previously reported as being induced by PMK-1 and negatively controlled by DAF-16 [32] . We examined the regulation of these genes during C . albicans infection and found that they were equally downregulated by C . albicans in both wild-type and daf-16 ( mgDf47 ) mutants ( Figure S2 ) , which suggests that DAF-16 is not responsible for this phenotype . In support of this observation , DAF-16::GFP remained localized to the cytoplasm following exposure to C . albicans and did not translocate into the nucleus , as it does when it is activated to regulate transcription ( data not shown ) . We also wondered whether signaling through the PMK-1 pathway results in the downregulation of antibacterial immune effectors during C . albicans infection . However , the basal expression of clec-67 and C17H12 . 8 was profoundly affected by PMK-1 ( Figure S2 ) , which precluded analysis of differential regulation during C . albicans infection in pmk-1 ( km25 ) mutants . In summary , we show that antibacterial response genes are downregulated during C . albicans infection , including a group whose basal expression is repressed by DAF-16 and stimulated by PMK-1 . We conclude that an unidentified mechanism , independent of DAF-16 , accounts for this phenotype .
Using a C . elegans pathogenesis assay that is conducted on solid agar plates , we show that C . albicans yeast cells kill worms in a manner dependent on live organisms and cause pathogenic distention of the nematode intestine during infection . Furthermore , we found that both heat-killed and virulence-attenuated C . albicans readily enter the nematode intestine , but are less pathogenic than wild-type yeast . While the mechanism of nematode mortality during C . albicans infection is unknown , these data suggest that some aspect of fungal virulence is required for yeast to infect and kill C . elegans . In response to C . albicans attack , we found that the nematode mounts a pathogen-specific defense response that involves the induction of antifungal effectors and core immune genes . Interestingly , 56% of the genes involved in the transcriptional response to C . albicans infection were also differentially regulated by heat-killed C . albicans . These data suggest that a large part of the transcriptional response to C . albicans is elicited by fungal PAMPs . In mammals , heat-killed fungi also strongly activate host defenses and have been used to study PAMP-mediated immune signaling [13] , [56] . In myeloid cells , cell wall components of heat-killed yeast ( mannans and β-glucans ) activate the pattern recognition receptors TLR2 , TLR4 , MR and dectin-1 to initiate antifungal immune responses [15] . Indeed , the process of heat killing may actually exaggerate innate immune responses in human cells by exposing fungal PAMPs . For example , β-glucans within the cell wall of C . albicans are normally covered by mannoproteins and thus blocked from detection by dectin-1 [13] , [51] . Treatment of yeast cells with heat depletes this protective layer and exposes β-glucans , thereby enhancing dectin-1-mediated proinflammatory cytokine responses [56] , [57] . The transcriptome profiling experiments and the expression analyses of nematodes infected with virulence-attenuated C . albicans suggest that factors associated with fungal virulence also elicit a transcriptional response in C . elegans . We do not know , however , whether these factors are derived from the host ( e . g . as a consequence of cell damage ) or from the pathogen . Recently , Moyes et al . found that human epithelial cells integrate inputs from C . albicans PAMPs via pattern recognition receptors together with “danger signals” perceived by the host during invasive fungal growth [14] . Interestingly , these researchers observed a biphasic activation of the p38 MAP kinase ( MAPK ) pathway , which was initially dependent on PAMP recognition and later on fungal burden and hyphal formation during invasive growth . We found a requirement for PMK-1 , the nematode ortholog of the p38 MAP kinase , in the response to C . albicans infection . We therefore propose that similar mechanisms of pathogen detection involving the PMK-1 pathway exist in C . elegans . As in the human epithelium , the nematode may integrate signals from PAMPs together with inputs associated with fungal virulence to delineate a “pattern of pathogenesis [58]” specific to fungal infection . Further research is needed to determine the PAMPs that are detected by C . elegans , the intestinal pattern recognition receptors that bind them and the mechanisms by which fungal virulence is perceived in the nematode . The immune response induced by Gram-negative bacteria , Gram-positive bacteria and fungi involve a small number of overlapping genes , a result that is somewhat surprising given the marked difference between prokaryotic and eukaryotic pathogens . Although others have also reported that the nematode mounts shared responses against different kinds of pathogens [34] , [36] , [59] , our data are the first to define a core set of immune regulators involved in the defense against three prototypical nosocomial pathogens . These findings may ultimately have clinical implications . Our laboratories and others are using C . elegans pathogenesis assays as a means to identify novel antimicrobial therapies with immunomodulatory activity [60] . Thus , identifying compounds that boost these core immune response genes may yield novel therapies that can cure infection by three diverse , nosocomial pathogens and may be a strategy that can be applied in higher order hosts . Unexpectedly , C . albicans infection of the nematode caused the downregulation of a number of antibacterial response genes including CUB-like genes , C-type lectins and ShK toxins . Moreover , it seems that both heat-killed ( non-pathogenic ) C . albicans and live ( infectious ) C . albicans can cause this repression . Interestingly , the basal expression of many of these genes is positively regulated by the p38 MAP kinase homolog PMK-1 and negatively regulated by DAF-16 . How might the selective downregulation of these antibacterial response genes be evolutionarily advantageous for the worm ? We know that the DAF-2 insulin/insulin-like growth factor receptor signals to the FOXO/forkhead transcription factor DAF-16 to control life span and stress resistance [61]–[63] and that DAF-16 negatively regulates P . aeruginosa immune response genes [32] . Troemel et al . postulated that immune response genes may be energetically expensive to make and thus their downregulation by DAF-16 under normal growth conditions may partially account for the lifespan-enhancing effects of DAF-2/DAF-16 pathway [32] . Irazoqui et al . found that the coordinated regulation of the immune response genes clec-60/61 and clec-70/71 influenced nematode survival . C . elegans animals carrying multiple copies of these gene clusters , which are induced during S . aureus infection , but not by P . aeruginosa or C . albicans , were more resistant to S . aureus , but were paradoxically hypersusceptible to P . aeruginosa [34] . We therefore propose that the transcriptional repression of antibacterial response genes , such as clec-60 and clec-70 , during C . albicans infection is an adaptive response . Given the recognized ability of FOXO/forkhead transcription factors to repress immune response genes both in C . elegans and in mammals [64] , we hypothesized that DAF-16 activity would be responsible for this phenotype . However , our data suggest that an unidentified mechanism , independent of DAF-16 , represses these genes following C . albicans infection . We are not aware of other examples in metazoans in which activation of specific antimicrobial defenses results in the transcriptional downregulation of another immune response . In contrast , this phenomenon is well described in the immune response of Arabidopsis thaliana , a widely-studied , model laboratory plant [65] . In Arabidopsis , as well as other plants , two low molecular weight immune hormones , salicylic acid and jasmonic acid , are involved in the activation of distinct immune response pathways . Salicylic acid is primarily activated by obligate , biotrophic pathogens that require living plant cells to acquire nutrients . Jasmonic acid , on the other hand , is involved in the response to necrotrophic pathogens that kill host cells and then feed on the carcasses . In most cases , activation of salicylic acid-mediated signaling downregulates jasmonic acid signaling and vice versa . The mutual antagonism of the salicylic acid and jasmonic acid pathways is generally interpreted in terms of evolutionary tradeoffs between biotrophic and necrotrophic defenses [65] . Our data suggest that a similar antagonism may be occurring in C . elegans between bacterial and fungal defenses . That is , when confronted with a virulent fungal pathogen , C . elegans focuses its immune response on the production of specific antifungal effectors at the expense of antibacterial defenses . Our analysis of the genes downregulated by P . aeruginosa or S . aureus did not reveal a statistically significant overlap with the genes induced following exposure to C . albicans . An alternative explanation is that the genes that are downregulated by C . albicans actually encode key immune effectors important for defense against both bacterial and fungal pathogens . Instead of the host downregulating the expression of these genes , the transcriptional repression may reflect an offensive measure by C . albicans to enhance its ability to infect C . elegans . In this study , we describe a novel C . elegans assay for the study of C . albicans yeast-mediated pathogenesis , which complements our hyphal formation model that we used to identify novel virulence determinants in C . albicans [33] . In our previous study , we screened a C . albicans mutant library containing homozygous mutations in 83 transcription factors [66] for clones attenuated both in their ability to form hyphae in vivo and kill C . elegans [33] . We uncovered several novel mediators of hyphal growth and showed that the efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant [8] , which is unable to program filamentation , was also attenuated for virulence in the C . elegans model , as it was in mammalian systems . The efg1Δ/efg1Δ cph1Δ/cph1Δ double mutant contain lesions in transcription factors that are the conserved readouts of the cAMP-mediated cascade ( Efg1p ) and the MAP-kinase cascade ( Cph1p ) , each with well-described roles in the control of morphogeneis and virulence [8] , [67] . In the current study , we show that this mutant was also attenuated for virulence in the C . elegans yeast-mediated pathogenesis assay . These data suggest that the C . albicans cAMP-mediated and MAP-kinase cascades also regulate yeast-specific virulence determinants and support the hypothesis that this morphogenic form is an important contributor to the pathogenic potential of wild-type fungi , as has been suggested by others [11] , [68]–[70] . These data also indicate that the C . elegans system can be used in large-scale screens of C . albicans mutant libraries for novel virulence regulators possessed by yeast .
C . elegans were maintained and propagated on E . coli OP50 as described [71] . The C . elegans strains used in this study were: N2 bristol [71] , pmk-1 ( km25 ) [22] , daf-16 ( mgDf47 ) [72] , fer-15 ( b26 ) ;fem-1 ( hc17 ) [55] , AU0157 [agEx39 ( myo-2::cherry , clec-60::GFP ) ] [28] and TJ356 [zIs356 ( pDAF-16::DAF-16-GFP;rol-6 ) ] [73] . The C . albicans strains used in this study were DAY185 ( ura3Δ::λimm434/ura3Δ::λimm434 ARG4:URA3::arg4::hisG/arg4::hisG his1::hisG::pHIS/his1::hisG ) [74] , SC5314 ( clinical isolate ) [75] and Can34 ( ura3Δ::λimm434/ura3Δ::λimm434 cph1Δ::hisG/cph1Δ::hisG efg1Δ::hisG/efg1Δ::hisG-URA3-hisG ) [8] . Unless otherwise specified , C . albicans DAY185 was used as the wild-type strain . Yeast strains were grown in liquid yeast extract-peptone-dextrose ( YPD , BD ) broth or on brain heart infusion agar containing 45 µg of kanamycin/ml at 30°C . Bacteria were grown in Luria Broth ( LB , BD ) . The previously described protocol for pathogen infection of C . elegans was modified for these studies [76] . Freshly grown C . albicans of the indicated genotype were picked from a single colony and used to inoculate 1 mL of YPD broth , which was allowed to grow overnight with agitation at 30°C . The following day , 10 µL of yeast were spread into a square lawn in a 4 cm tissue culture plate ( BD ) containing 4 mL of BHI agar and kanamycin ( 45 µg/mL ) . For experiments that compared heat-killed and live C . albicans , cells were subjected to the exact same preparatory conditions . A single colony of yeast was grown in 1 mL BHI at 30°C overnight and then inoculated into 50 mL YPD . After approximately 20 hours of incubation , cells were split into two aliquots , collected by centrifugation and washed twice with sterile PBS ( pH 7 . 4 ) . One aliquot was resuspended in 1 mL PBS , exposed to 75°C for 60 minutes and washed again with sterile PBS . The other aliquot was processed in parallel with the heat-killed sample . Cells were suspended in 25 mL PBS , incubated at room temperature for 60 minutes and washed again with sterile PBS . 10 µL of this sample were added to the killing assay plates . To heat kill E . coli , a similar protocol was followed except that a single colony was inoculated into 50 mL LB and allowed to grow overnight at 37°C . Cells were exposed to 75°C for 30 minutes . In both cases , heat-killed organisms were plated on YPD or LB agar to ensure no viable organisms remained . 50 µL of heat-killed cells were added to the assay plates . The plates were then incubated for approximately 20 hours at 30°C . The next day , a Pasteur pipette molded into the shape of hockey stick was used to gently scrape excess yeast off the top of the thick C . albicans lawn . This step greatly facilitated scoring the animals as live or dead on subsequent days and did not affect the pathogenicity of C . albicans ( data not shown ) . Five-fluoro-2′-deoxyuridine ( FUDR; 75–100 µg/mL ) was added to the plates 1 to 2 hours before the start of the assay to reduce the growth of progeny and prevent matricidal killing of nematodes by C . albicans . Thirty to forty young adult animals of the indicated genotype were added to each of three assay plates per condition studied . Although it is possible that microorganism inocula varied among individual worms , we doubt that such variation affected the pathogenicity of C . albicans in our assay since we observed similar killing kinetics in replicate experiments . Animals were scored as live or dead on a daily basis by gently touching them with a platinum wire . Worms that crawled onto the wall of the tissue culture plate were eliminated from the analysis . All killing assays were conducted at 25°C . C . elegans survival was examined using the Kaplan-Meier method and differences were determined with the log-rank test ( STATA 6; STATA , College Station , TX ) . N2 animals were synchronized by hypochlorite treatment . Arrested L1s were plated on 10 cm NGM plates seeded with E . coli OP50 and grown at 20°C until they were young adults . Animals were then added to 10 cm plates containing 20 mL of BHI agar ( with 45 µg of kanamycin/ml ) and live C . albicans , heat-killed C . albicans or heat-killed E . coli . Plates were prepared using the method described above except 50 µL of cells were added to the plates for each condition together with 200 µL of PBS to facilitate even dispersion of the microbes . Three separate biological replicates of nematodes were exposed to these conditions for 4 hours at 25°C . RNA was extracted using TRI Reagent ( Molecular Research Center ) according to the manufacturer's instructions and purified using an RNeasy column ( Qiagen ) . RNA samples were prepared and hybridized to Affymetrix full-genome GeneChips for C . elegans at the Harvard Medical School Biopolymer Facility following previously described protocols [32] and instructions from Affymetrix . Data were analyzed using Resolver Gene Expression Data Analysis System , version 5 . 1 ( Rosetta Inpharmatics ) . Three biologic replicates per condition were normalized using the Resolver intensity error model for single color chips [77] . Conditions were compared using Resolver to determine the fold change between conditions for each probe set and to generate a P value using a modified t-test . Probe sets were considered differentially expressed if the fold change was 2-fold or greater ( P<0 . 01 ) . When comparing datasets , the overlap expected by chance alone was determined in 50 groups of randomly selected C . elegans genes using Regulatory Sequence Analysis Tools ( http://rsat . ulb . ac . be/rsat/ ) , a technique that has been used for similar analyses [78] . P values were determined using chi-square tests . Analyses for over-representation of GO annotation categories were performed using DAVID Bioinformatics Resources 6 . 7 from the National Institute of Allergy and Infectious Diseases [79] , [80] . Two databases were used to determine the expression patterns for selected genes: Expression Patterns for C . elegans Promoter::GFP Fusions ( http://gfpweb . aecom . yu . edu/ ) [81] and NEXTDB [82] . Animals were treated and RNA was extracted as described above . RNA was reverse transcribed to cDNA using the Retroscript kit ( Ambion ) . cDNA was analyzed by qRT-PCR using a CFX1000 machine ( Bio-Rad ) and previously published primers [32] , [39] , [41] . Primer sequences for fipr-22/23 ( GCTGAAGCTCCACACATCC and TATCCCATTCCTCCGTATCC ) and cnc-7 ( CAGGTTCAATGCAGTATGGCTATGG and GGACGGTACATTCCCATACC ) were designed for this study , checked for specificity against the C . elegans genome and tested for efficiency with a dilution series of template . The primer set for fipr-22/23 cannot distinguish between these two genes owing to sequence similarity . All values were normalized against the control gene snb-1 , which has been used previously in qRT-PCR studies of C . elegans innate immunity [31] , [32] , [48] , [83] . Analysis of the microarray expression data revealed that the expression of snb-1 did not vary under the conditions tested in our experiment . Fold change was calculated using the Pfaffl method [84] and compared using t-tests . Nematodes were mounted onto agar pads , paralyzed with 10 mM levamisole ( Sigma ) and photographed using a Zeiss AXIO Imager Z1 microscope with a Zeiss AxioCam HRm camera and Axiovision 4 . 6 ( Zeiss ) software . Accession numbers for the genes and gene products mentioned in this paper are given for Wormbase , a publically available database that can be accessed at http://www . wormbase . org . These accession numbers are pmk-1 ( B0218 . 3 ) , abf-2 ( C50F2 . 10 ) , fipr-22/23 ( C37A5 . 2/4 ) , cnc-4 ( F09B5 . 9 ) , cnc-7 ( F53H2 . 2 ) , cht-1 ( C04F6 . 3 ) , T19H5 . 1 , irg-3 ( F53E10 . 4 ) , clec-67 ( F56D6 . 2 ) , K08D8 . 5 , C17H12 . 8 , F49F1 . 6 , F35E12 . 5 , F01D5 . 5 , clec-60 ( ZK666 . 6 ) and daf-16 ( R13H8 . 1 ) . The microarray dataset can be downloaded from the National Center for Biotechnology Gene Expression Omnibus ( GEO; http://www . ncbi . nlm . nih . gov/geo ) . The accession number for these data is GSE2740 . | Despite being a part of the normal flora of healthy individuals , Candida albicans is the most common fungal pathogen of humans and can cause infections that are associated with staggeringly high mortality rates . Here we devise a model for the study of the host immune response to C . albicans infection using the nematode C . elegans . We found that infection with the yeast form of C . albicans induces rapid and robust transcriptional changes in C . elegans . Analyses of these differentially regulated genes indicate that the nematode mounts antifungal defenses that are remarkably distinct from the host responses to pathogenic bacteria and that the nematode recognizes components possessed by heat-killed C . albicans to initiate this response . Interestingly , during infection with a pathogenic fungus , the nematode downregulates antibacterial immune response genes , which may reflect an evolutionary tradeoff between bacterial and fungal defense . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"immunology",
"microbiology",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"fungal",
"diseases",
"infectious",
"diseases",
"mycosis",
"biology",
"immunity",
"innate",
"immunity",
"yeast",
"and",
"fungal",
"models",
"candida",
"albic... | 2011 | Candida albicans Infection of Caenorhabditis elegans Induces Antifungal Immune Defenses |
Leptospirosis causes significant morbidity and mortality worldwide; however , the role of the host immune response in disease progression and high case fatality ( >10–50% ) is poorly understood . We conducted a multi-parameter investigation of patients with acute leptospirosis to identify mechanisms associated with case fatality . Whole blood transcriptional profiling of 16 hospitalized Brazilian patients with acute leptospirosis ( 13 survivors , 3 deceased ) revealed fatal cases had lower expression of the antimicrobial peptide , cathelicidin , and chemokines , but more abundant pro-inflammatory cytokine receptors . In contrast , survivors generated strong adaptive immune signatures , including transcripts relevant to antigen presentation and immunoglobulin production . In an independent cohort ( 23 survivors , 22 deceased ) , fatal cases had higher bacterial loads ( P = 0 . 0004 ) and lower anti-Leptospira antibody titers ( P = 0 . 02 ) at the time of hospitalization , independent of the duration of illness . Low serum cathelicidin and RANTES levels during acute illness were independent risk factors for higher bacterial loads ( P = 0 . 005 ) and death ( P = 0 . 04 ) , respectively . To investigate the mechanism of cathelicidin in patients surviving acute disease , we administered LL-37 , the active peptide of cathelicidin , in a hamster model of lethal leptospirosis and found it significantly decreased bacterial loads and increased survival . Our findings indicate that the host immune response plays a central role in severe leptospirosis disease progression . While drawn from a limited study size , significant conclusions include that poor clinical outcomes are associated with high systemic bacterial loads , and a decreased antibody response . Furthermore , our data identified a key role for the antimicrobial peptide , cathelicidin , in mounting an effective bactericidal response against the pathogen , which represents a valuable new therapeutic approach for leptospirosis .
Pathogenic Leptospira spp cause life-threatening disease , primarily in the world’s most impoverished populations [1] . Leptospirosis is considered the most widespread zoonotic disease due to the large number of wild and domestic mammalian reservoirs [2] and causes an estimated 1 . 03 million infections and 59 , 000 deaths globally per year [3 , 4] . In Brazil alone , epidemic outbreaks of leptospirosis in urban slum communities during seasonal periods of heavy rainfall account for more than 10 , 000 reported cases each year [5 , 6] . Despite its widespread importance , development of a vaccine has been hampered by genetic and antigenic diversity in pathogenic Leptospira , which are comprised of ten species and >200 serovars . Humans are accidental hosts and acquire the disease through contact with water or soil contaminated with Leptospira excreted in the urine of reservoir hosts . During a systemic infection , clinical manifestations can range from a self-limiting febrile illness to Weil´s disease , the classic severe form with jaundice , acute renal failure and bleeding , or severe pulmonary hemorrhage syndrome ( LPHS ) [1 , 7 , 8] . Notably , case fatality rates from Weil’s disease and LPHS are >10% and 50% , respectively [7 , 8 , 9 , 10] . At present , the factors contributing to disease progression and poor clinical outcomes in patients with leptospirosis are poorly understood . No studies to date have found associations between genetic differences in Leptospira spp and poor disease outcomes , suggesting other factors drive disease severity [11 , 12] . The infecting inoculum dose may also affect patient outcomes , but these have been intrinsically difficult to measure and evaluate . Alternatively , differences in host factors , such as the immune response to bacteria , are known to contribute in general to the development of lung injury and septic shock , and may be relevant to severity of responses to Leptospira infection [13–16] . Several lines of evidence suggest that the pathology associated with severe disease , LPHS and Weil’s syndrome , is in part , immune-mediated . In the city of Salvador , Brazil , a single serovar , L . interrogans serovar Copenhageni , causes the full spectrum of disease , suggesting that strain-specific differences in pathogen virulence do not explain differences in disease outcome [7 , 17–19] . Furthermore , patients with poor outcomes , such as fatality , have been shown to have altered cytokine responses , including elevated mRNA transcripts of IL-1α and its antagonist receptor , IL-1RA , higher serum levels of IL-10 and IL-6 , and high ratios of IL-10:TNFα [20–24] . These cytokines are commonly associated with innate immune responses; however , such cytokine responses are largely uncharacterized in patients with leptospirosis , despite neutrophilia being a common disease characteristic , and the known protective or detrimental roles neutrophils play in other bacterial infections [25–27] . Potential roles for T cells and endothelial cells in poor disease outcomes have also been described , but these remain less well validated in patient investigations [28] . While antibodies appear protective in experimental animal models of leptospirosis [29–32] , definitive roles for B and T cells in the resolution or exacerbation of human Leptospira infections remain largely uncharacterized . A better understanding of the human response to Leptospira infection could discern likely pathogenic processes involved in disease development . To identify features of disease response associated with death or survival , we conducted an in-depth multi-parameter analysis of immune responses during the acute phase of leptospirosis in a well-characterized cohort of hospitalized patients , including assessment of transcriptional profiles , serum components , and immune cell abundances . This work contributes to our understanding of immunopathogenic processes that affect disease outcome and identifies novel approaches to therapeutic intervention for leptospirosis .
To identify host factors contributing to fatality , we enrolled 16 patients hospitalized with acute leptospirosis ( 13 survivors , 3 fatal cases ) and 4 healthy community volunteers for in-depth characterization of clinical course and immune responses . Table 1 describes the patient characteristics for biochemical and clinical values during hospitalization for fatal and nonfatal cases . As noted in other studies , we observed that fatal cases had significantly elevated percentages of neutrophils as well as lower minimum hematocrit and percent lymphocytes in peripheral blood [23 , 26] . We also found that acute phase anti-Leptospira agglutinating antibody titers were lower and Leptospira loads trended higher in the deceased group . Of the outcomes measured , we determined that only acute lung injury was more frequently associated with the deceased group . We found no differences in days of symptoms prior to admission ( P = 0 . 26 ) , age ( P = 0 . 42 ) , gender ( P = 0 . 35 ) , or days of symptoms prior to microarray sampling ( P = 0 . 33 ) between survivors ( 8 . 4 ± 1 . 9 days ) and nonsurvivors ( 6 . 7 ± 2 . 3 days ) . Thus , our patient cohort is representative of disease outcomes common to leptospirosis in Brazilian patients . To delineate host responses important during acute leptospirosis , we performed a transcriptional analysis of whole blood comparing paired samples from acute disease and convalescence from 13 survivors and a single sample from four healthy Brazilian volunteers ( S1 Text ) . As expected , many genes ( 1089 unique transcripts ) were differentially expressed during acute infection relative to convalescence ( S1A Fig ) . Of these , 363 transcripts increased in relative abundance during acute illness relative to convalescence , while 726 increased in convalescence relative to acute infection ( S1 Table ) . To identify pathways associated with the acute phase of leptospirosis , we performed a functional analysis ( DAVID ) of all 1089 significant transcripts and found 40 significant ( FDR < 0 . 01 and Benjamini <0 . 05 ) Gene Ontology terms ( GO terms ) enriched in acute versus convalescent comparisons ( S2 Table ) [33 , 34] . These include categories such as “response to bacterium” , “defense response” , “antigen binding” , and 72 transcripts for immunoglobulin or immunoglobulin-like genes , which were enriched 2 . 0–7 . 9-fold during acute illness . Notably , no genes were significantly different between the convalescent and healthy volunteer groups , indicating that the immune state had returned to baseline 1–3 months following hospitalization ( S1 Fig ) . When we examined patterns of gene expression in the 16 cases using principal component analysis , the first principal component ( PC1 ) ( which explained 9 . 9% of the variance ) , separated the acute patient samples from paired convalescent and healthy volunteer samples ( P = 0 . 0001 ) ( Fig 1A ) . Strikingly , we found fatal acute disease profiles separated significantly from survivor profiles in principal component 2 ( PC2 ) ( 7 . 6% variance; P = 0 . 014 ) . These data support the hypothesis that host-derived factors are associated with fatal outcomes . We therefore directly compared the acute phase transcriptional profiles of 13 nonfatal and three fatal cases to identify specific gene expression changes associated with survival and death . We identified 389 differentially expressed ( DE ) unique transcripts in deceased patients versus survivors ( Fig 1B ) . We categorized the DE transcripts into three expression profile groups based on co-expression patterns after hierarchical clustering ( Figs 1B and S1A ) . Groups 1 and 2 represent transcripts more abundant in nonfatal cases , with group 2 transcripts ( 92 unique genes ) elevated during the acute phase of illness compared to convalescence , and Group 1 transcripts ( 76 unique genes ) stable across nonfatal cases and not significantly different from convalescence ( Fig 1B and S3 Table ) . Group 3 contains 221 transcripts with higher abundance in deceased patients compared to acute phase survivors or convalescents ( Fig 1B and S3 Table ) . Despite survivors presenting with varying disease severity during acute infection , only 27% ( N = 105 ) of all significant transcripts from acute phase survivors ( compared to convalescence ) exhibited differential expression when compared with those of deceased patients ( S1D Fig and S1 Table ) . Further , a majority of all the transcripts , elevated during acute infection in survivors , were not elevated during acute infection in deceased patients , suggesting a specific transcriptional alteration in fatal cases ( S1D Fig ) . We performed functional enrichment analyses for transcripts more abundant in each of the three deceased vs survivor signature groups to discover molecular mechanisms that may have contributed to fatal disease outcomes ( Fig 1C–1E; S2 and S4 Tables ) [33 , 34] . Within Group 1 transcripts , we identified 38 significant GO terms and 30 REACTOME pathways ( Fig 1C; S2 and S4 Tables ) , the vast majority of which were related to immune function or coagulation . Of note , the chemokine CCL5 ( RANTES ) , important for recruitment of T cells , leukocytes and NK cells , had 4 . 3-fold lower expression in fatal cases ( Fig 2 ) . We observed similar reductions ( 2 . 6–3 . 0-fold ) in three chemokine receptor transcripts , CX3CR1 , CXCR3 , and CCR3 ( Fig 2 ) . Fatal cases also had 2 . 0–5 . 0-fold lower abundance of six genes involved in blood coagulation , most notably platelet factor 4 ( PF4/CXCL4 ) , pro-platelet basic protein ( PPBP/CXCL7 ) , and Factor 13 ( F13A1 ) ( Fig 2 ) . Together these data suggest that fatal cases had diminished migration of immune cells to sites of infection as well as reduced expression of coagulation factors , which could contribute to the hemorrhaging observed in many fatal leptospirosis cases . We identified a prominent diminution in the abundance of Groups 1 and 2 transcripts involved in antigen presentation and the generation of an adaptive immune response in fatal cases ( Figs 1C , 1D and 2; S2 and S4 Tables ) including 2 . 1–3 . 9-fold reductions in the abundance of six HLA Class II transcripts and CD74 ( invariant chain ) . We observed reduced abundance of 10 transcripts involved in T cell activation and regulation in fatal cases such as 2 . 7 and 3 . 2-fold decreased abundance of CD40LG , a T cell protein , which promotes immunoglobulin class switching , and CD27 , important for T and B cell memory and immunoglobulin class switching ( Fig 2 ) . Further , we identified decreased abundance of 25 pathways related to B cell and antibody responses in fatal cases ( S2 and S4 Tables ) , with a 2 . 8- to 13 . 6-fold decreased expression for 47 immunoglobulin genes and reduced abundance of transcripts for B cell signaling ( BLNK ) , IgM production ( MZB1 ) , and germinal center formation ( POU2AF10 ) ( Fig 2 ) . These results suggest that fatal cases may not be capable of mounting robust T cell and B cell responses during acute infection because of defects in antigen presentation . Adult patients with Gram negative septic shock also generate transcriptional profiles with reduced T cell activation and antigen presentation suggesting fatal leptospirosis cases may share clinical features with bacterial sepsis [35] . To examine whether patients with severe infection had diminished adaptive immune cell activation or frequencies , we employed multi-parameter flow cytometry of peripheral blood mononuclear cells ( PBMCs ) to profile T cell and B cell responses in 11/13 survivors and 1/3 deaths ( Fig 3 ) . Acute lung injury ( ALI , defined in Materials and Methods ) is a significant risk factor for death in leptospirosis [8 , 17] . Because we had limited PBMCs from deceased patients , we stratified patients by ALI to distinguish cases with higher probabilities of fatality . The ALI group had significantly fewer CD4+ and CD8+ T cells and larger percentages of naïve B cells [36] . In contrast , the patient group lacking pulmonary complications ( No ALI ) had elevated memory B cell and transitional B cell populations , the subsets required for antibody production [36] . Both the B and T cell subsets associated with immune activation and memory were lower in the more severe ALI group , which included one fatal case ( Fig 3 ) [36] . These phenotypic changes are consistent with our microarray findings and suggest that fatal cases had dampened adaptive immune responses in the peripheral blood . Because we observed a significant reduction in transcription of immunoglobulin-encoding genes in fatal cases and reduced B and T cell responses in more severe disease , we quantified anti-Leptospira agglutinating antibodies in corresponding sera from the 16 patients with microarray results and an additional 18 fatal cases ( N = 21 total ) and 11 survivors ( N = 24 ) . Notably , we found that anti-Leptospira antibody titers were significantly lower in fatal cases ( Tables 1 and S6 ) . This is consistent with a decreased abundance of immunoglobulin transcripts ( Fig 1B–1D ) . We also identified a significant correlation between levels of transcription of 21 immunoglobulin genes and agglutinating antibody titers during early acute infections , indicating a direct association between transcript levels and antibody titers ( S5 Table ) . Further , the higher systemic bacterial loads detected in fatal cases inversely correlated with both immunoglobulin gene transcripts and antibody titers ( β = -0 . 3811 ± 0 . 1554 , P = 0 . 0188 ) , providing functional data suggesting a critical role for decreased humoral responses in fatal cases . The transcript with the greatest difference in abundance ( 17 . 6-fold ) between nonfatal and fatal cases encodes an antimicrobial peptide ( AMP ) , cathelicidin ( CAMP ) ( Fig 2 and S3 Table ) . In survivors , cathelicidin had 20 . 3-fold higher expression during acute disease compared to convalescence ( S1 Table ) . Interestingly , we found no association between disease outcome and other antimicrobial molecules produced by innate immune cells such as resistins , defensins , or elastase , although we detected increased abundance of these transcripts in acute illness relative to convalescence in survivors ( S1 Table ) . Therefore , cathelicidin is the only antimicrobial peptide with significantly decreased expression in fatal cases . In addition to cathelicidin , fatal cases had many transcripts with significantly elevated expression compared to survivors ( Group 3; 221/389 ) , including two GO terms , “Interleukin 1 Receptor Activity” and “Sulfur Compound Biosynthetic Processes” , and two related functional pathways “IL-1 Signaling” and “Metabolism” ( S2 and S4 Tables ) . Concordantly , we measured large relative increases in expression ( 2 . 9–9 . 4-fold ) for the decoy IL-1 receptor ( IL-1R2 ) , the IL-1 receptor ( IL-1R1 ) , and IL-18 receptor ( IL-18R ) , indicating transcription of these proinflammatory pathways may be relevant to outcome in fatal cases ( Fig 2 ) . We identified increased abundance of transcripts involved in NF-κB signaling , a pathway important for proinflammatory responses: MKK3 and MKK6 , members of p38 signaling pathways that respond to environmental stress . Additionally , we found elevated transcript levels of human growth factor ( HGF ) , a gene induced by proinflammatory cytokines , although this may be in response to signaling or driven by higher bacterial loads [37 , 38] . Together , these data suggest that the increased abundance of specific proinflammatory responses in nonsurvivors may have contributed to fatality . The transcriptional studies identified more than 30 genes with striking differences between survivors and non-survivors , which may shed light on pathogenesis or have potential as new therapeutic targets or diagnostic markers ( Fig 2 ) . To investigate some of these targets , we quantitated serum levels of LL-37 ( active fragment of cathelicidin , CAMP ) , IL-18 , RANTES , HGF , and CHI3L1 by single or multiplex ELISA . ( Fig 4A–4F; N = 45 patients , 22 of whom died during acute infection ) . Notably , we measured significantly higher serum levels of LL-37 in survivors , consistent with microarray findings , while finding no differences in the levels of elastase , an enzyme produced by neutrophils , suggesting some normal neutrophil function . Consistent with our microarrays , we found elevated serum protein levels of RANTES in survivors , and lower levels of HGF and IL-18 , the ligand for IL-18R . These data provide further evidence that these genes and their products may play critical roles in disease progression . Lastly , levels of CHI3L1 protein were lower in survivors than in fatal cases ( Fig 4 ) . We do not know why these findings for CHI3L1 contrast with the results of gene expression analyses; however , lower CHI3L1 in surviving patients is consistent with its presence as a biomarker of severity in other inflammatory diseases [39] . To identify factors associated with case fatality including the clinical , transcriptional , cell subset , and serum factors assessed in stratified leptospirosis patients , we employed univariate analyses of data for the entire patient cohort ( S6 Table ) . As we noted in our initial cohort assessed for transcriptional analysis ( Table 1 ) , fatal cases had lower platelet counts , lower antibody titers , and higher bacterial loads . We also found an association of ALI with death . In contrast , survivors had less evidence of renal failure as measured by significantly lower maximum blood urea nitrogen , creatinine , and potassium levels , fewer hemodialysis treatments , and lower incidence of oliguria or anuria ( S6 Table ) . To identify independent risk factors for higher bacterial loads and death , we included all significant univariate variables ( Fig 4 ) and days of symptoms in multivariate linear and logistic regression models , respectively ( S1 Text ) . These analyses revealed that survivors had significantly higher titers of agglutinating antibodies ( β = -0 . 3811 ± 0 . 1554; P = 0 . 02 ) , and further , that lower serum levels of cathelicidin ( LL-37 ) predicted higher bacterial loads ( Table 2 ) . Additionally , lower RANTES levels and higher CHI3L1 serum levels were independent risk factors for death in patients with leptospirosis ( Table 2 ) . Gender ( P = 0 . 17 ) , age ( P = 0 . 10 ) , and days of symptoms ( P = 0 . 09 ) , possible confounders of disease outcome , were not significantly different between the two patient groups . As our results suggest a critical role for cathelicidin during infection with Leptospira spp , we tested the effect of LL-37 , the active peptide of cathelicidin , in a hamster model of lethal leptospirosis . Immediately prior to lethal infection with 100 live Leptospira interrogans serovar Copenhageni , we injected hamsters with LL-37 reconstituted in ddH2O , an LL-37 scrambled peptide reconstituted in ddH2O ( control group ) , or water alone ( control group ) . We found that while all hamsters in both control groups ( N = 14 ) died within 11 days of infection with high blood bacteremia , hamsters treated with LL-37 ( N = 7 ) were significantly protected from lethal infection , and controlled systemic bacterial loads ( Figs 5 and S3 ) . These data provide strong evidence that cathelicidin is a critical immune molecule protecting against fatal leptospirosis .
Despite the important global disease burden of leptospirosis [3 , 4] , there are key gaps in our understanding of host pathogenic mechanisms that contribute to poor disease outcomes such as massive pulmonary hemorrhage and death . To identify host factors contributing to fatality , we conducted an in-depth characterization of clinical , transcriptional , immune cell subset , and serum factors in hospitalized leptospirosis patients , including the first comprehensive human transcriptome analysis of peripheral blood during acute leptospirosis . We demonstrated that low serum levels of cathelicidin ( LL-37 ) is a risk factor for high bacterial loads and suggests cathelicidin is a novel , potential therapeutic for leptospirosis . Additionally , we identified CHI3L1 and RANTES , as new risk factors for death from leptospirosis . Our data suggests a lower magnitude of specific innate immune responses may underlie poor early control of infection and diminished activation of adaptive immune responses . Subsequently , increased bacterial proliferation promotes systemic inflammation , contributing eventually to patient death . The mechanistic details of this proposed model of pathogenesis remain to be determined . The most pronounced finding in the transcriptional profiling was the markedly lower level of transcripts encoding the antimicrobial peptide , cathelicidin , in fatal cases . The defect in production of antimicrobial peptides was not a global innate immune dysfunction , as we found no significant differences in other antimicrobial transcripts or serum proteins ( elastase , resistins , and defensins ) between survivors and fatal cases . We identified differences in abundances in only two Toll-like Receptors: TLR8 ( can detect single-stranded bacterial RNA ) [40 , 41] , which was elevated in fatal cases likely due to higher bacterial loads , and TLR7 ( senses bacterial RNA in phagosomes ) [42 , 43] , which was less abundant in fatal cases , possibly due to fewer phagocytic cells . Cathelicidin functions as an antimicrobial peptide , capable of directly killing bacteria , fungi , parasites , and some viruses [37] . Consistent with our results , direct anti-leptospiricidal activity has been demonstrated for the active peptide of cathelicidin , LL-37 , in vitro [44 , 45] . Unlike other antimicrobial peptides , cathelicidin is also an important activator of neutrophils , stimulating phagocytosis , diminishing apoptosis , and reducing LPS-driven TLR-dependent proinflammatory responses [37] . Reduced levels of circulating cathelicidin therefore could contribute to elevated bacterial load , which we observed in the hamster model , and higher levels of proinflammatory cytokines , such as IL-1 and IL-18 , which we observed in fatal human cases . Consistent with our current findings , others and we have shown previously that high levels of proinflammatory cytokines , and their transcripts , IL-1α , IL-6 , and IL-8 as well as the IL-1 antagonist receptor 1 , are associated with poor disease outcomes for leptospirosis [21 , 23 , 24 , 46] . These results strongly suggest that decreased cathelicidin might contribute both to decreased bactericidal activity and increased levels of inflammation , resulting in greater tissue damage and higher bacterial loads . These findings , combined with our animal experiments and recent biochemical [47] and clinical studies [48 , 49] involving cathelicidin , suggest a potential role for cathelicidin during acute illness as a novel therapeutic option for patients with leptospirosis . We identified several markers of inflammation in fatal cases: CHI3L1 , HGF , and proinflammatory cytokine receptors , IL-18R and IL-1R1 . CHI3L1 expression is induced by proinflammatory cytokines , and is associated with increased patient mortality in sepsis and other infectious or inflammatory diseases [50] . Proinflammatory cytokines also induce expression of HGF , a pleiotropic cytokine , which decreases inflammation , inhibits antigen presentation , and promotes organ injury repair [38] . HGF promoted healing in a mouse model of lung injury , and is in early clinical trials for reducing inflammation in acute spinal cord injuries [38] . We detected higher levels of HGF in fatal cases , suggesting these patients had greater systemic inflammation than survivors . IL-1 and IL-18 are cytokines produced following TLR signaling and inflammasome activation to induce downstream immune responses and inflammation [51] . Patients with poor disease outcomes from other critical illness , such as sepsis , also have elevated levels of IL-18 [13 , 52 , 53] . Several clinical trials are assessing the efficacy of IL-18 inhibition in primarily chronic inflammatory diseases , but their application to leptospirosis will require consideration of potential protective roles for IL-18 . Together , these data suggest CHI3L1 , IL-18 , and HGF represent new potential prognostic and therapeutic strategies for leptospirosis . Our study illustrated the importance of the adaptive immune response , and in particular the antibody response , in protection from fatal leptospirosis . While the humoral immune response is accepted widely as the primary mode of immunity to Leptospira infection , a protective role for antibodies has not been demonstrated definitively in humans . Passive transfer experiments in animal models of leptospirosis have shown that anti-LPS antibodies confer protection from homologous reinfection [54 , 55] . In keeping with these data , we detected significantly lower antibody titers and transcript abundance for immunoglobulins in patients that did not survive . The notable decrease in chemokines , such as RANTES , which functions to recruit immune cells to sites of infection , and which we identified as a risk factor for death , suggests aberrant cell trafficking could contribute to poor or slower adaptive immune response generation in fatal cases . However , further studies are needed to determine the mechanistic causes of neutropenia and lymphocytopenia in fatal cases , despite lower LL-37 and chemokine levels . Lastly , we observed a larger number of memory B cell and transitional B cell responses in patients with less severe leptospirosis , raising the intriguing idea that more severe disease may represent primary infection and that secondary infections , where some memory B cell responses are available for recall , may be less severe . Taken together , our data support the animal data in which anti-Leptospira antibodies are critical for bacterial clearance and improved disease outcomes [12 , 56 , 57] . The associations we identified in our microarray findings are strengthened by the functional assays we performed on the larger cohort of confirmed patients and the animal studies . However , our patients represent primarily individuals of mixed Caucasian and African descent and it will be important to identify whether the pathways we identified are generalizable to global populations , given that several studies have shown association of specific alleles with increased susceptibility to leptospirosis [15 , 16 , 58 , 59] . Further , it will be important to compare our findings on whole blood transcriptional profile with samples from the lungs in patients that develop LPHS . Studies of the specific tissue site may reveal additional immune dysfunction in the lungs . Our study provides the first evidence that patients die from leptospirosis because of a failure to mount innate and adaptive immune response to this pathogen . While we were able to analyze only a small number of patients , the results demonstrate the power of using systems biology approaches to understand disease pathology . We have identified several unique targets , which may represent new diagnostic and treatment of leptospirosis patients at greatest risk of death . CHI3L1 and RANTES serum levels are attractive candidate diagnostic markers , which could identify patients at risk for developing severe disease and allow hospitals to focus limited resources on patients with greatest risk . Most importantly , the development of anti-Leptospira antibody therapies or administration of cathelicidin are potential new strategies for reducing bacterial loads in severely ill patients .
The Yale Institutional Review Board ( HIC#1006006956 ) , the Ethics Committees at Fiocruz-Salvador ( CEP-CPqGM 329 ) and Hospital Couto Maia ( 175 ) , and the Brazilian Ministry of Health National Ethics Committee in Research ( CONEP 15925 ) approved the study protocol prior to study initiation . Our trained study team obtained written informed consent in the native language ( Portuguese ) from all participants prior to blood and data collection . All animal protocols and work were approved and conducted under the guidelines of the Yale Institutional Animal Care and Use Committee ( IACUC ) , under approved protocol #2014–11424 . The Yale IACUC strictly adheres to all Federal and State regulations , including the Animal Welfare Act , those specified by Public Health Service , and the US Department of Agriculture , and uses the US Government Principles for the Utilization and Care of Vertebrate Animals Used in Testing , Research , and Training as a guide for all animal studies . We performed active surveillance at an infectious disease hospital in Salvador , Brazil , to identify patients with suspected leptospirosis between April 2013 and September 2013 with the goal of discovering markers associated with case fatalities . We used previously described criteria to identify cases: <15 days of fever , jaundice , high serum creatinine and/or blood urea nitrogen , acute lung injury ( [ALI]; defined by mechanical ventilation , ≥250 mL blood in lungs or endotracheal tube , and/or respiration rate >38/min ) , oliguria ( <500 mL urine/24 h ) , and epidemiologic data supporting likelihood of exposure to Leptospira spp [17] . For transcription studies , we stratified patients by survival , and for immunophenotyping by ALI [17] . We confirmed cases using serum microagglutination test ( MAT ) ( 13/16 ) , qPCR ( Leptospira genome/mL blood ) ( 5/16 ) , and/or blood culture ( 2/16 ) , as described previously [7 , 17 , 18 , 60 , 61] . We collected clinical data during patient interviews and from hospital charts for all enrolled patients using a RedCap database [62 , 63] . In surviving patients , we collected two venous blood samples: acute phase ( ≤72h of hospital admission; one patient collected at 168h; mean collection time: survivors 8 . 4 ±1 . 9d , fatal cases 6 . 7±2 . 3d ) and convalescence ( 32-90d post-admission ) . We collected the identical acute sample from fatal cases , and a sample from four healthy individuals with prior Leptospira exposure ( 303-367d post-admission ) . We collected whole blood directly into red-top tubes ( sera for ELISAs , MATs , and MSDs ) , PAXgene solution ( RNA microarrays ) , CPT tubes ( peripheral blood mononuclear cells [PBMCs] ) , EDTA tubes ( qPCR ) , or EMJH culture medium ( blood culture ) , processed and froze all samples at -70°C the same day of collection . We bar-coded all samples , monitored transport temperature , and recorded all cold chain data including sample receipt , processing time , and freezing time . We isolated peripheral blood mononuclear cells ( PBMCs ) from the blood of leptospirosis patients using CPT tubes and cryopreserved them in 90% FBS containing 10% DMSO and stored in liquid nitrogen until batch analysis as described [70] . On the day of analysis , we thawed cells and labeled them with fluorescent antibodies for immunophenotyping as follows: 1 ) T cell panel: HLA-DR , CD38 , CD28 , CD8 , CCR7 , CD45RA , CD27 , and CD4; 2 ) TH1/2/17 cell panel: CD4 , CD38 , CD45RO , CD8 , CXCR3 , CCR6 , CXCR5 , and CCR4; 3 ) Treg cell panel: HLA-DR , CD127 , Foxp3 , CD45RO , CD25 , CCR4 , CD39 , and CD4; and 4 ) B cell panel: IgD , CD38 , CD20 , CD24 , CD27 , and CD10 [36] . We analyzed cells by flow cytometry using a custom , programmed BioMek robotic platform and detected using an LSR Fortessa ( BD BioSciences ) [70] . We employed two-dimensional gating analysis of flow cytometry files by FlowJo ( Treestar ) to remove doublets and debris using scatter channels . We labeled living cells with a viability marker and pre-gated for T cells ( CD3+ ) or B cells ( CD3- ) . Immunophenotyping panels defined T cell subsets ( TH1/2/17 cell , and Treg ) or B cells ( CD3–CD19+ ) . We clustered cell subsets as defined above using Citrus version 0 . 08 ( https://github . com/nolanlab/citrus ) to compare no ALI and ALI ( met criteria for ALI described above and/or died ) samples [71] . The SAM model type employed file sample size of 200 events , and the minimum cluster size was <5% , significance for false discovery rate ( FDR ) ( q < 0 . 05 ) . We performed each comparison at least 3 times to ensure reproducibility [71] . We quantified the levels of LL-37 , the active peptide form of cathelicidin ( HyCult Biotech , Cat#HK321-02 ) , and elastase ( Hycult Biotech , Cat#HK319 ) by ELISA using duplicate dilutions of sera collected from the patients described in this study , and sera frozen at -80°C from an additional 33 patients ( 49 total ) with laboratory-confirmed leptospirosis: 13 survivors ( 25 total ) and 21 nonsurvivors ( 24 total ) . Due to sera availability , we measured elastase in only 29 patients: 14 survivors and 15 deceased patients . We measured serum levels of IL-18 , CHI3L1 , HGF , and CCL5 using technical replicates on single-plex MSD kits for each molecule as specified by the manufacturer ( Meso Scale Discovery , IL-18: K151MCD-1; RANTES Ultra-Sensitive Kit: K151BFC-1; HGF: K151HDC-1; and CHI3L1/YKL-40: K151NHD-1 ) . We intraperitoneally infected 3-week old Golden Syrian hamsters with 100 live leptospires ( Leptospira interrogans serovar Copenhageni strain Fiocruz L1-130 ) immediately following intracardiac injection of 1 mg/kg LL-37 ( synthetic peptide ) in ddH2O ( BACHEM; treated group ) , 1 mg/kg scrambled LL-37 ( scrambled control [BACHEM Cat . H-7886] ) , or the identical volume of ddH2O ( ddH2O control group ) [72 , 73] . On days 4 , 6 , and 8 after infection , we performed qPCR on peripheral blood as described above . We monitored animals a minimum of two times daily . We immediately euthanized moribund or animals with signs of clinical disease by CO2 inhalation . We used GraphPad Prism 6 . 0 , R , and EpiInfo 7 to perform all statistical analyses except for microarray data , which we analyzed as described above . We performed descriptive statistics on continuous variables , and used the Fisher exact test or Mann-Whitney t-test to compare categorical or continuous variables , respectively , between survivor and deceased groups . We performed linear regression and logistic regressions in R , using backward elimination , to predict bacterial load and death , respectively . For the multivariate regression predicting death , we used a backward elimination approach to identify the best model fit using variables that were significantly associated with death in univariate analysis and days of symptoms prior to blood collection . We did not include HGF in the logistic regression analysis due to a high number of outliers resulting in non-linearity of features ( S2 Fig ) . We considered P<0 . 05 significant . | Leptospirosis causes over one million cases and nearly 60 , 000 deaths annually . Infection with the spirochetal bacterium results in a spectrum of symptoms , ranging from mild febrile illness to life-threatening pulmonary hemorrhage syndrome and acute kidney injury . Despite leptospirosis being a leading cause of zoonotic morbidity worldwide , little is known about the human immune response to Leptospira infections , and less about the pathogenic mechanisms resulting in severe disease outcomes . Here , we used a systems biology approach to discover transcripts and immunoprofiles associated with case fatality . We identified new risk factors for high bacterial loads and fatal leptospirosis , including the antimicrobial peptide , cathelicidin , which we validated in an animal model . Cathelicidin therefore represents a potential novel treatment for severe cases of leptospirosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"immune",
"physiology",
"immune",
"cells",
"pathogens",
"immunology",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"animals... | 2016 | Cathelicidin Insufficiency in Patients with Fatal Leptospirosis |
The functional separation of ON and OFF pathways , one of the fundamental features of the visual system , starts in the retina . During postnatal development , some retinal ganglion cells ( RGCs ) whose dendrites arborize in both ON and OFF sublaminae of the inner plexiform layer transform into RGCs with dendrites that monostratify in either the ON or OFF sublamina , acquiring final dendritic morphology in a subtype-dependent manner . Little is known about how the receptive field ( RF ) properties of ON , OFF , and ON-OFF RGCs mature during this time because of the lack of a reliable and efficient method to classify RGCs into these subtypes . To address this deficiency , we developed an innovative variant of Spike Triggered Covariance ( STC ) analysis , which we term Spike Triggered Covariance – Non-Centered ( STC-NC ) analysis . Using a multi-electrode array ( MEA ) , we recorded the responses of a large population of mouse RGCs to a Gaussian white noise stimulus . As expected , the Spike-Triggered Average ( STA ) fails to identify responses driven by symmetric static nonlinearities such as those that underlie ON-OFF center RGC behavior . The STC-NC technique , in contrast , provides an efficient means to identify ON-OFF responses and quantify their RF center sizes accurately . Using this new tool , we find that RGCs gradually develop sensitivity to focal stimulation after eye opening , that the percentage of ON-OFF center cells decreases with age , and that RF centers of ON and ON-OFF cells become smaller . Importantly , we demonstrate for the first time that neurotrophin-3 ( NT-3 ) regulates the development of physiological properties of ON-OFF center RGCs . Overexpression of NT-3 leads to the precocious maturation of RGC responsiveness and accelerates the developmental decrease of RF center size in ON-OFF cells . In summary , our study introduces STC-NC analysis which successfully identifies subtype RGCs and demonstrates how RF development relates to a neurotrophic driver in the retina .
Many studies have investigated the segregation of ON and OFF pathways in the retina during postnatal development , and much is known about the structural maturation of different subtypes of retinal ganglion cells ( RGCs ) [1] , [2] . For example , based upon the sublamina in which RGC dendrites arborize in the inner plexiform layer ( IPL ) , RGCs can be classified into three subtypes: ON , OFF , and ON-OFF , which presumably respond to light onset , light offset , and both [3]–[5] . RGCs acquire their final dendritic branching pattern and territories in a subtype-dependent manner [6]–[9] . In the mouse , RGC dendritic arbors ramify diffusely in the IPL shortly after birth and then undergo extensive laminar refinement [8]–[11] . Consequently , the fraction of bistratified RGCs decreases as they are converted into monostratified cells during the first postnatal month [8] , [10] . While RGCs having dendrites monostratified in the ON sublamina continue to expand their dendritic field size by adding new branches following eye-opening , bistratified RGCs acquire their final morphology by the time of eye opening [8] , [9] . Far less is known about how the development of the physiological properties of different RGC subtypes might correlate with their dendritic refinement during postnatal development . This is largely due to the lack of a reliable method to identify ON , OFF and ON-OFF center RGCs in the mouse . The full field flash stimulus is often used in visual experiments [11]–[13]; for example , Tian and Copenhagen ( 2003 ) showed that with this stimulus the number of RGCs with ON-OFF responses decreases after eye-opening . However , because full field flashes stimulate both the center and the surround of the receptive field ( RF ) , responses evoked by this stimulus cannot be linked reliably to center-type . Furthermore , RF structure cannot be studied with full field flashes because of the spatially uniform nature of the stimulus . Spatiotemporal white noise [14] has become a quite commonly used high-dimensional visual stimulus to investigate the spatial extent and temporal properties of RFs . Visual neural responses to white noise are typically modeled with a linear filter followed by a static nonlinearity ( the LN model ) [15] , [16] . The Spike-Triggered Average ( STA ) , which is the mean visual stimulus that precedes a spike , can be calculated from a cell's response to white noise , and is often used as the linear filter in LN models of ON and OFF center RGCs [16] . The STA approach struggles however when responses depend on symmetric static nonlinearities [17] , such as might be expected for an RGC with an ON-OFF center . For example , the STA technique has been applied to classify RGCs into different ON or OFF subtypes , but not the ON-OFF subtype [18] . In such situations , spike-triggered covariance ( STC ) analysis , which identifies multiple relevant linear filters , provides a better analytical approach [19]–[21] . However , full STC analysis can be cumbersome and data-hungry , therefore we develop in the current study a Non-Centered Spike-Triggered Covariance ( STC-NC ) analysis which maintains the simplicity of a single filter analysis but is capable of characterizing different RGC subtypes , including the ON-OFF center variety . The molecular players involved in RGC dendritic development have begun to be identified . Both brain-derived neurotrophic factor ( BDNF ) and Neurotrophin-3 ( NT-3 ) have been shown to modulate RGC laminar refinement and dendritic branching in a subtype-specific manner [8] , [9] . We and others have shown that both brain-derived neurotrophic factor ( BDNF ) and Neurotrophin-3 ( NT-3 ) modulate RGC laminar refinement and dendritic branching during postnatal development [8]–[10] . While both BDNF and NT-3 accelerate RGC laminar refinement , their effects on dendritic branching is cell-type specific [8] , [9] . BDNF selectively promotes formation of new branches in monostratified , but not bistratified RGCs [8] . By contrast , NT-3 promotes the formation of more , but shorter , branches selectively in bistratified RGCs , resulting in smaller but more dense dendritic trees [9] . It is less clear how these neurotrophin-dependent dendritic structural changes of RGCs relate to refinement of function . Here we have investigated the physiological development of RGC RF properties after eye-opening , and the regulation of this process by NT-3 . We introduce and characterize the STC-NC analysis , which is an innovative variant of the STC method and able to accurately identify ON , OFF , and ON-OFF center character in RGCs . With this new method , we show that wildtype RGCs gradually develop sensitivity to focal stimulation after eye-opening . The percentage of ON-OFF center cells decreases with age , and the RF centers of ON and ON-OFF cells become smaller . Importantly , overexpression of NT-3 leads to a precocious maturation of RGC responsiveness and accelerates the developmental decrease of RF center size in ON-OFF cells .
Three types of visual stimuli , the full field flash , a spot stimulus , and a white-noise checkerboard stimulus , were applied to investigate RF properties of RGCs ( Fig . 1 ) . The full field flash stimulus ( Fig . 1A ) is easy to generate , and with high contrast it evokes robust responses [11] , [22] . Classification through the Response Dominance Index ( RDI , see Materials and Methods ) identifies positive values with ON and negative values with OFF RGCs . Those cells which give robust responses and have an RDI near zero are conventionally classified as ON-OFF RGCs ( Fig . 1F ) . However , as noted earlier , full field flash stimuli drive both center and surround RF components , making classification based on responses to this stimulus an inaccurate reflection of RGC center type . By contrast , a localized flashing spot only stimulates the RF center of an individual cell ( Fig . 1B ) , and the Spot Response Bias can be calculated to identify ON , OFF and ON-OFF centers ( Fig . 1F–G ) . However , this approach provides a time-consuming and inefficient way to identify RGCs when a microelectrode array system is used where the discharges of many cells can be recorded ( and potentially stimulated ) simultaneously . Compared to the full field flash and the spot stimulus , a white-noise checkerboard stimulus ( Fig . 1C ) permits focal stimulation of many RGC centers simultaneously . It thus combines the spatial localization advantage of spot stimulation with the advantage enjoyed by the full field flash stimulus of gathering data from a large population of cells at once . The STA is the standard measure employed for identifying RF center responses with checkerboard stimuli . As expected , STAs possessed large deviations from the mean luminance ( 2 cd m−2 ) only in a small spatially confined region of the display , the center of the neuron's RF , and tapered to near mean luminance away from the RF center , indicating that stimulus perturbations at these distant locations were uncorrelated with spikes ( Fig . 1D ) . To determine the center size , a bivariate Gaussian was fit to the single STA frame of maximal contrast and the area within the ellipse formed by the fitted Gaussian's 1σ contour used ( Fig . 1D ) . For each frame of the STA , pixel contrast was averaged over the spatial STA indices located within the 1σ contour of the bivariate Gaussian , and these mean contrast intensities were plotted as a function of time prior to the spike ( Fig . 1E ) . This time course , which is biphasic in structure for most cells and monophasic for a small number of cells , provides identification of ON and OFF character . However , a large population of bistratifying presumed ON-OFF center cells are present in the mouse retina [8] , [11] , [23] , so some cells identified as ON or OFF center with the STA could well be ON-OFF center cells with unbalanced ON and OFF components . We therefore labeled cells either ON-dominating ( instead of ON-center , see below ) if the maximal positive contrast preceded the spike more closely than the maximal negative contrast , or OFF-dominating in the converse case . We compared cell identification with the full field flash RDI , Spot Response Bias and the STA . We find that the full field flash RDI correlates positively with Spot Response Bias ( R = 0 . 73 , p<0 . 001 ) , but in many cases the RDI failed to reveal the character of the RF center determined by the spot stimulus ( Fig . 1F ) . RGCs with RDI values near zero , suggesting an ON-OFF response , ranged in Spot Bias from 1 ( ON-center ) to −1 ( OFF-center , Fig . 1F ) . Similarly , correlation between STA values and Spot Response Bias is unsatisfactory; RGCs with both ON and OFF STA signatures exhibited a broad range of Spot Response Bias values ( Fig . 1G ) . Cells possessing ON-OFF center character as determined by the spot stimulus were included in both ON and OFF STA classes ( Fig . 1G ) . Clearly , there were many instances where the full field flash RDI and the STA measures failed to provide an accurate , or even coherent , classification of RGC cell-type , particularly for cells with ON-OFF character ( also see Table S1 for a summary ) . To provide a more accurate characterization of RGC subtypes , including the ON-OFF center type , we employed white noise analysis techniques that were based on the second moment of the spike-triggered ensemble ( STE ) , rather than upon its mean ( i . e . STA ) . STC analysis proceeds by performing a principal component analysis ( PCA ) on the STE . PCA can be achieved by eigendecomposition of the covariance matrix , which generates eigenvectors that are then sorted by their eigenvalues to identify directions of large and small variance . Traditionally , PCA is performed by the eigendecomposition of the mean-centered covariance matrix . Mean centering is mathematically founded as it guarantees that the low-dimensional hyperplane created by the few principal components with greatest eigenvalues is the best fitting hyperplane for the high-dimensional data set in the mean square error sense [24] . Additionally , mean centering ensures that during dimensionality reduction by projection of the data onto the top principal components , maximal data variance is retained [24] . Because the retinal circuitry is divided into ON and OFF pathways , the polarity of the stimulus that is required to elicit a spike is of primary importance . The stimulus polarity is determined by the positive or negative deviation of the stimulus from mean luminance . For this reason , mean luminance , which is represented by the zero vector in our high dimensional stimulus space , is a critical reference point . To identify the single direction in stimulus space that maximizes the second moment of the STE about zero , we chose to perform eigendecomposition on a non-centered second moment matrix M: In the case of a non-centered moment matrix , the eigenvector with the greatest eigenvalue maximizes the second moment of the STE around zero – not the variance , which is the second moment around the mean . The hyperplane created by the eigenvectors with the greatest eigenvalues is the best-fitting hyperplane that passes through the origin [24] , [25] . We term this technique STC-NC analysis . We compared the results of our new technique with those of conventional mean-centered STC analysis in Fig . 2 ( A–D ) . These panels plot the full set of the STE , utilizing a geometrical representation of an M-dimensional stimulus space , where M is the number of pixel intensities in a spatiotemporal stimulus . In this space , each individual spatiotemporal stimulus , which is simply a vector of pixel intensities , can be represented by a single point . Choice of axes is particularly important because they are used to reduce the dimensions of the space for a more manageable presentation and analysis . The perpendicular axes of the space were determined by the conventional STC analysis , which was performed either with ( Fig . 2A–B ) or without ( Fig . 2C–D ) projecting the STA out of the STE [21] , [26] . The red vector illustrates the direction of the STC-NC and its relationship to the STA and the STC principal components . The length of the STC-NC vector , compared to the length of the axes , is used to graphically represent the degree to which the M-dimensional vector projects into the 2D space shown . For cells possessing strong ON-OFF character , STC analysis captured the most important stimulus dimension in the highest variance PC , when the STA was not projected out of the STE ( Fig . 2D ) . The STC-NC closely matched the low variance PC ( STC500 ) for cells of ON or OFF character ( ) , and matched the high variance PC ( STC1 ) for cells of strong ON-OFF character ( , Fig . 2D ) . In this case , the STA was near orthogonal to this direction ( , Fig . 2D ) , showing that this measure captured little of this response-related variance . As a comparison , cells with ON or OFF character were often best described by the lowest variance PC . This PC aligned extremely well with the STA , as evidenced by the near unity normalized inner product of the STA with the low variance PC ( ) . A similar pattern was seen when the STA was projected out of the STE prior to analysis . While the STC-NC very closely aligns with the STA for ON or OFF cells ( , Fig . 2A ) , the STC-NC again clocked toward the high variance PC to better capture the structure for ON-OFF cells ( Fig . 2B ) . However , the STC-NC did not as closely match the STC1 axis as it did in Fig . 2D . The separation observed between the STC-NC and the STC1 axes in Fig . 2B demonstrates that projecting out the STA can significantly interfere with the discovery of the direction of highest variance . Taken together , our data demonstrate that the STC-NC can be used alone to consistently and reliably identify the stimulus feature most relevant to the classification of the cell as ON , OFF , or ON-OFF . By contrast , with conventional STC , the most important stimulus dimension is captured by different directions in the analysis , depending upon the character of the cell and whether or not the STA is projected out of the STE ( Fig . 2A–D ) . This property complicates STC analysis , and requires careful inspection of the STE along the several dimensions defined by the STA , the low , and the high variance STC PCs in order to identify cell-type . In addition , the STC-NC linear filter strongly aligns with the largest or smallest eigenvecter of the STC analysis ( Fig . 2C–D ) , thus the predictive capabilities of the STC-NC analysis are essentially the same as the predictive capabilities of a single-filter STC analysis . By contrast , the conventional STC analysis allows for a larger number of filters , and , obviously , with extra parameters to describe the data , it can generally be expected to provide better predictive capabilities . However , the benefit obtained by utilizing a larger number of STC filters is highly dependent upon the cell's RF properties . When ON and OFF subfields of the ON-OFF center cells have similar locations and temporal properties , the advantage of extra filters is small . Moreover , each additional filter places an additional load on the amount of data needed to characterize it and adds to the complexity of the analysis . Overall , the simplicity offered by the STC-NC analysis constitutes an important advantage for the STC-NC over conventional STC analysis ( Table S1 compares the different methods ) . Because the STC-NC identifies the single direction of greatest deviation from mean luminance in the STE , the underlying filter strength is high , and the technique is not overly data-hungry . We computed the STC-NC at 100-spike intervals for a subset of cells to study convergence using a 500-dimensional stimulus ( Fig . S1 ) . We considered an STC-NC vector to be fully converged only if the second half of the unit's spikes caused the projection of the estimated vector onto the final vector to change by <5% . In these converged cells , for 90% of the length of the estimated STC-NC vector to project onto the final STC-NC , 2600±300 spikes were needed , and for 80% of the length to project onto the final STC-NC vector , only 1400±150 spikes were required ( n = 10 , Fig . S1 ) . We further show that the LN model generated with the STC-NC more accurately reflects the underlying mechanisms of the neuron than the model generated with the STA ( Fig . 2E–F ) . We reconstructed the LN models for the stimulus features that are represented by the directions in the high-dimensional stimulus space of panel B ( Fig . 2E–F ) . The RF was described by the linear component of the model ( Fig . 2E–F , top ) . The STC-NC consistently revealed a more structured spatial RF for ON-OFF cells than the STA did ( Fig . 2E–F , top ) . Although the RFs of both the STA and the STC-NC frequently overlapped in space , their forms were different: the STA often demonstrated an unstructured bipolar RF ( Fig . 2E , top ) while the STC-NC demonstrated a well defined unipolar RF ( Fig . 2F , top ) . Next , using the projection of the STE onto each linear filter ( Fig . 2E–F , middle ) , the corresponding static nonlinearities were recovered ( Fig . 2E–F , bottom ) . Only the STC-NC revealed a symmetric static nonlinearity reflective of the ON-OFF character of the cell ( Fig . 2F ) . In the LN model presented above , the STC-NC used a single filter to describe an ON-OFF center cell , with the implicit assumption that the converging ON and OFF RF centers were spatiotemporally overlapping and inverted . Gollisch and Meister ( 2008 ) demonstrated a technique by which the differing ON and OFF linear filters can be identified by computing the STA for a single cell's separated ON and OFF responses ( STA-ON and STA-OFF; Fig . 2G–H ) . Utilizing this technique in conjunction with STC-NC analysis , which directly separated ON-type and OFF-type spikes , we observed that the ON-OFF cells possessed well-overlapped ON and OFF spatial RF components ( Fig . 2G–H; insets ) . But while some ON-OFF cells possessed similar STA-ON and STA-OFF temporal filters ( Fig . 2G ) , others possessed slower STA-OFF filters ( Fig . 2H ) . Interestingly , the temporal mismatch between the ON and OFF RF components often created a triphasic STA ( data not shown ) [17] . Thus , while the STC-NC accurately separates the ON and OFF spikes from an ON-OFF cell , the LN model generated with this technique does not always fully describe the dynamics of the component ON or OFF pathways . The STC-NC Bias ( see Materials and Methods ) correlated well with the Spot Response Bias ( Fig . 3 ) . ON cells identified by the Spot Response Bias ( Fig . 3A ) had unimodal 1D STE projections onto the STC-NC , and these distributions were shifted toward positive outputs from the linear filter , creating an asymmetric static nonlinearity and a positive STC-NC Bias ( Fig . 3B ) . ON-OFF cells identified by the Spot Response Bias ( Fig . 3C ) generally possessed bimodal 1D STE projections creating a more symmetric static nonlinearity ( Fig . 3D ) and an STC-NC Bias closer to zero . STC-NC Bias values corresponded well with Spot Response Bias values ( R = 0 . 84 , p<0 . 001 , Fig . 3E ) , indicating that STC-NC analysis and responses to spot stimuli generally predict the same kind of RF center behavior . To further test our classification of cells into ON , OFF , or ON-OFF subtypes using STC-NC , we examined the STC-NC Bias against a measure of bimodality . Using the Hartigan and Hartigan Dip Test , a p value was obtained for the null hypothesis that the 1D STE comes from a unimodal distribution ( Fig . 4A ) . As expected from initial observations , cells formed three distinct clusters . OFF cells were unimodal with large negative STC-NC bias ( Fig . 4B#1 ) , ON cells were unimodal with large positive STC-NC bias ( Fig . 4B#3 ) , and ON-OFF cells were bimodal with STC-NC bias closer to zero ( Fig . 4B#2 ) . The strongly bimodal character of the ON-OFF cells is reassuring , as it indicates the presence of easily separable ON and OFF RF components . We chose to define cells with an STC-NC Bias <−0 . 6 as OFF cells , cells with an STC-NC Bias >0 . 6 as ON cells , and those with intermediate values as ON-OFF cells . While most cells fitted nicely into these three classes , we noted that a small number of cells did not . Most cells with bimodal 1D STE distributions had a relatively symmetric static nonlinearity ( Fig . 4B#2 ) , but some were unbalanced ( Fig . 4C#4 ) . These cells possessed an STC-NC bias far from zero and were presumably not ON-OFF centered . Additionally , while the majority of cells with STC-NC bias near zero had bimodal 1D STE distributions , a few cells were determined to have unimodal distributions ( Fig . 4C#5 ) . The 1D STE distributions of many of these cells departed substantially from the normal distribution , suggesting mutimodality , but no clear secondary modes were evident . The 1D STE projections of these cells resembled the projections of previously described “ring cells” [19] , suggesting that they may be sensitive to additional stimulus directions in the M dimensional stimulus space . These exceptions were comparatively rare in our cell sample and have minimal impact on the classification of cell-types in the data that follow . Using the newly-developed STC-NC method , we characterized development of RF properties of RGCs in wildtype ( WT ) and NT-3 overexpression mice . WT retinas had poor light responses immediately following eye opening ( Fig . 5A–B ) . At P15 , fewer cells had above-threshold STC-NCs ( 26 cells/retina , n = 6 ) than at P18 ( 81 cells/retina , n = 4; p = 0 . 001 in Student's t-test; Fig . 5A ) . The number of cells per retina responding to the visual stimulus at P18 was not significantly different from P25 ( 51 cells/retina , n = 4; p = 0 . 11 in Student's t-test; Fig . 5A ) . In addition , we also noted that a large percentage of spike trains recorded at P15 lacked a defined RF and these cells were unclassifiable ( Fig . S2A ) . About 35% of spike trains were discarded for this reason at P15 , significantly higher than those at P18 ( P = 0 . 04 in Student's t-test , Fig . S2A ) . When the strength of an RGC's visual response was quantified by normalizing the STC-NC's greatest absolute contrast with the standard deviation of its surrounding spatiotemporal elements , we found that the population-averaged STC-NC signal strength was greatest at P18 ( p<0 . 0001 in One-way ANOVA test , Fig . 5B ) , indicating a decline in responsiveness from P18 to P25 . As demonstrated earlier , the STA analysis does not identify the RF properties of ON-OFF RGCs properly . However , as the STA represented the existing state of the art , we examined the cell number and signal strength using the STA analysis to provide a baseline against which to compare the results obtained using STC-NC analysis . Similarly , fewer cells had above-threshold STAs at P15 ( 45 cells/retina , n = 6 ) than at P18 ( 88 cells/retina , n = 4 ) and at P25 ( 84 cells/retina , n = 4; p<0 . 01 in Student's t-test; Fig . 5A ) and the population-averaged STA signal strength was greatest at P18 ( p<0 . 001 in One-way ANOVA test , Fig . 5B ) . In NT-3 overexpression ( OE ) mice , cells exhibit more mature light-response characteristics than WT controls at P15 . First , there were more responsive cells in NT-3 OE retinas ( 62 cells/retina , n = 4 ) than WT at P15 ( 26 cells/retina , n = 6; Fig . 5A ) , though it did not reach statistical significance with our sample size ( p = 0 . 1 in Student's t-test ) . By P18 and P25 , the number of responsive cells in NT-3 OE mice was not significantly different from WT controls ( p = 0 . 5 , Fig . 5A ) . Secondly , the population-averaged STC-NC signal strength was higher in NT-3 OE mice than in WT at P15 ( p<0 . 0001 ) and at P18 ( p<0 . 0001 , Fig . 5B ) . The difference disappeared by P25 ( p = 0 . 5 , Fig . 5B ) . Similar results were obtained from the STA analysis that NT-3 OE mice tended to have more responsive cells at P15 ( 45 cells/retina , n = 6 , p = 0 . 1 ) and stronger signal at both P15 ( p<0 . 001 ) and P18 ( p<0 . 01 , Fig . 5B ) . We confirmed that NT-3 OE mice had normal RGC morphology at P16 compared to WT controls ( Fig . 5C–D ) . RGC were visualized by staining with Brn-3a , a transcription factor for subtype RGCs , and SMI-32 , a marker for neurofilaments in subtype RGCs [9] , in whole-mounted retinas at P16 ( Fig . 5C ) . Brn-3a immuno-labels the nuclei of subtype RGCs , and SMI-32 immuno-labels the ganglion cell bodies ( Fig . 5C ) . Overexpression of NT-3 did not affect cell density of Brn-3a positive RGCs ( WT: n = 6 retinas; NT-3 OE: n = 4; P = 0 . 14 in Student's t-test ) , nor did it alter the cell body size of SMI-32 positive RGCs ( WT: n = 4; NT-3 OE: n = 3; P = 0 . 18 , Fig . 5D ) . Together , these data strongly indicate that overexpression of NT-3 accelerates the development of RGC light responses to focal stimuli after eye opening . We next examined the developmental change of the ON-OFF RGC population , which can now be accurately identified by the new STC-NC method . Previous work showed that cells with bi-stratified dendritic trees ( presumptive ON-OFF cells ) were gradually converted into cells with mono-stratified dendritic trees ( presumptive ON or OFF cells ) after eye opening [8] , . We therefore quantified changes in the percentage of different RGC subtypes in the developing WT retina using STC-NC analysis , to examine whether the RGC dendritic laminar refinement correlates with the development of the RF-center properties of different RGC subtypes . Because WT retinas at P15 yielded a smaller number of recorded cells per retina ( Fig . 5A ) , possessed a larger percentage of unmappable , visually unresponsive cells ( Fig . S2A ) , and demonstrated lower average STC-NC and STA signal response strengths ( Fig . 5A ) , we focused the STC-NC classification analysis on data collected from P18 and P25 mice . Cumulative histograms of the absolute value of the STC-NC Bias demonstrated that during development , the percentage of cells with strong ON-OFF character was reduced at P25 compared to P18 ( p = 0 . 003 in K-S Test , Fig . 6A ) . In addition , we found a significant difference in the RGC subtype composition of WT retinas at P18 and P25 ( Fig . 6B , two sample χ2 p = 0 . 015 ) . At P25 , the percentage of ON-OFF cells was reduced compared to P18 , and the percentage of ON cells was increased correspondingly ( Fig . 6B ) . The percentage of OFF cells was similar for the two ages ( Fig . 6B ) . Our data demonstrate therefore that the percentage of ON-OFF center RGCs decreases while the percentage of ON center RGCs increases with age after eye opening . We compared WT to NT-3 OE retinas at P18 and found no significant changes in the percentage of the three RGC subtypes in NT-3 OE retinas ( Fig . 6C , two sample χ2 p = 0 . 366 ) . Prior work has shown that NT-3 OE mice had fewer cells with bi-laminated dendritic trees at P13 compared to age-matched WTs , however , this difference disappeared by P28 [9] . In this study , we also found , consistent with the anatomical results , that the distribution of the three physiological RGC subtypes was the same in NT-3 OE retinas and WT controls at P25 ( data not shown ) . With the superior cell classification and improved RF mapping provided by the STC-NC , we examined the development of RF center sizes in WT retinas in a subtype-specific manner ( Fig . 7 ) . We found that ON-OFF cells exhibited a significant decrease in RF center size at P25 compared to P18 ( Fig . 7B ) . The mean RF center size for ON-OFF cells at P25 was 13 . 9±0 . 6×103 µm2 , significantly lower than that at P18 ( 15 . 6±0 . 4×103 µm2 , p = 0 . 03 in Wilcoxon rank sum; same below ) . The RF center sizes of ON cells also decreased from P18 to P25 ( P18: 14 . 3±0 . 3×103 µm2; P25: 13 . 3±0 . 4×103 µm2; p = 0 . 03; Fig . 7B ) . OFF cells displayed a slight trend towards smaller RF center size too , but this change was not statistically significant ( P18: 15 . 0±0 . 8×103 µm2; P25: 14 . 1±1 . 0×103 µm2; p = 0 . 64 , Fig . 7B ) . These data demonstrate that the center size of ON and ON-OFF cells decrease with age , while the size of OFF-cell centers is unchanged . To rule out the possibility that a smaller RF center was due to a weaker center response , we plotted the STC-NC signal strength against RF size for WT ON cells at ages of P18 and P25 ( Fig . 7C ) . We found that the RF sizes demonstrated weak but negative correlations with their signal strength ( P18: R = −0 . 34 , P≤10−4; P25: P25: R = −0 . 11 , P = 0 . 23; Fig . 7C ) . Our data suggest that increased signal strength is associated with smaller , but not larger , RF size . To isolate the significance of the group variable alone , we thus utilized the Analysis of Covariance technique ( ANCOVA ) to compensate for variations in STC-NC signal strength . We still found that P25 ON cells and ON-OFF cells were significantly smaller ( ON: p = 4×10−5; ON-OFF: p = 5×10−3 ) , and that OFF cells were unchanged ( p = 0 . 68 in ANCOVA test ) . For P15 WT animals , we were able to characterize , on average , only 26 cells per retina with STC-NC analysis . Dividing this number among the three classes , we might expect to record roughly 13 ON cells , 10 ON-OFF cells , and only 3 OFF cells ( Fig . S2B ) . Nevertheless , we were able to collect data from enough samples to analyze their RF center sizes ( Fig . S2C–F ) . We found that the RF sizes for OFF and ON-OFF cells increased from P15 to P18 ( OFF: n = 23 cells , p = 0 . 04; ON-OFF: n = 41 cells , p = 0 . 003 in Wilcoxon rank sum , Fig . S2C–F ) . For ON cells , the change was not significant ( n = 91 cells , p = 0 . 09 , Fig . S2D ) . Taken together , our data suggest that different subtypes of RGCs exhibit different growth patterns after eye opening and cannot be therefore passive processes resulting simply from growth of the eye . We also confirmed that with STC-NC a single bivariate Gaussian function fitted the RF center better that it did for the STA when considered over all cell-types ( Fig . S3 ) . For ON and OFF cells , both the STA and the STC-NC accurately described the spatial RF structure , and as expected , the STA and the STC-NC gave similar RF center sizes for these RGC subtypes ( Fig . S3 ) . For ON-OFF cells , however , the STA failed to characterize the RF center ( Fig . 2B , E ) ; the unstructured center given by the STA did not fit a bivariate Gaussian well and the resulting estimate of center size was poor . The STA measurement of RF center size for ON-OFF cells was significantly smaller than the STC-NC measurement ( p<0 . 001 , Fig . S3 ) . The STA therefore poorly estimates ON-OFF cell RF center size — a problem remedied by the use of STC-NC analysis . We next investigated whether NT-3 regulates the RF center size of different RGC subtypes after eye opening ( Fig . 8 ) . With the STC-NC , we found that ON-OFF cells in NT-3 OE mice exhibited significantly smaller RF centers compared to WT retinas at P18 ( WT: 15 . 6±0 . 4×103 µm2; n = 118; NT-3 OE: 13 . 7±0 . 3×103 µm2 , n = 170; p<0 . 001 in Wilcoxon rank sum , Fig . 8B–C ) . ON cells also possessed smaller center sizes in NT-3 OE mice at P18 , although this difference was less pronounced than it was for ON-OFF cells ( WT: 14 . 3±0 . 3×103 µm2; n = 169; NT-3 OE: 13 . 5±0 . 3×103 µm2 , n = 208; p = 0 . 04; Fig . 8A–B ) . Although the RF center sizes of OFF cells tended to be smaller in NT-3 OE mice compared to WT the difference did not reach statistical significance with our sample size ( WT: 15 . 0±0 . 9×103 µm2; n = 40; NT-3 OE: 13 . 7±0 . 5×103 µm2 , n = 72; p = 0 . 24 , Fig . 8A–B ) . At P25 , ON-OFF cells continue to have significantly smaller RF centers in NT-3 OE mice ( Fig . 8E–F ) . The mean center size of ON-OFF cells at P25 in NT-3 OE mice was 12 . 3±0 . 5×103 µm2 , significantly smaller than that of WT controls ( 13 . 9±0 . 6×103 µm2 , p = 0 . 04 in Wilcoxon rank sum , Fig . 8E–F ) . The small difference between WT and NT-3 OE mice in the center size of ON cells at P18 had disappeared by P25 , and what difference in center size there might have been for OFF cells at P18 was also absent by P25 ( ON: p = 0 . 83; OFF: p = 0 . 92 , Fig . 8E ) . With ANCOVA statistical test to compensate for the increases in STC-NC signal strength in NT-3 OE mice , we continued to find that at P18 NT-3 OE retinas had smaller ON-OFF cells than WT controls ( p = 0 . 03 ) , but not for ON and OFF cells ( ON: p = 0 . 28; OFF: p = 0 . 19 ) . This same pattern was observed at P25 , where ON-OFF cells in NT-3 OE retinas were again smaller than WT controls ( ON-OFF: p = 0 . 04; ON: p = 0 . 42; OFF: p = 0 . 85 in ANCOVA test ) . Previous work has shown that bistratified ON-OFF cells have smaller dendritic field sizes in NT-3 OE mice [9] . With the STC-NC analysis , we have shown here that overexpression of NT-3 accelerates the developmental decrease of RF center size in ON-OFF cells , consistent with the earlier anatomical work .
RGCs are often modeled with a linear filter followed by a static nonlinearity ( the LN model ) . The visual stimulus is convolved with the linear filter and the output passed through a static nonlinearity which controls the expectation of spike discharge [15] , [16] , [19] , [21] . The STA is often used as the linear filter in LN models of ON and OFF center RGCs [16] , [18] . Because the STA is the average stimulus preceding a spike , if the underlying nonlinear mechanism of an LN neuron is highly symmetric , as one might expect for ON-OFF cells , the STA is unable to recover the linear filter [21] . STC analysis is a technique used in combination with the STA to identify additional linear filters for neurons in the visual system [19]–[21] . A full STC analysis can be particularly powerful when used to uncover multiple filters from the complex RFs of neurons in the visual cortex [20] . When applied to retinal neurons , STC is also capable of identifying the filters used by ON-OFF RGCs ( Fig . 2 ) . But , when applied to the entire population of RGCs , the primary RF mechanism is not captured by a single STC filter . Instead , it may be described by any one ( or combination ) of the STA , the low , or the high variance eigenvectors , depending on the character of the RGC and whether or not the STC analysis is performed in a space perpendicular to the STA ( Fig . 2 ) . In this study , we demonstrated that the STC-NC greatly simplifies classification of RGCs and that it has an intuitive interpretation . Performing PCA on a non-centered moment matrix is justified , particularly when the zero vector ( mean luminance ) is an important reference [24] , [27] . Because this technique is not centered with the STA , it maximizes the second moment of the STE about mean luminance , not the variance ( which by definition is the second moment of the STE about its own mean ) . Using a non-centered second moment matrix allows us to find the single direction of maximal deviation from mean luminance regardless of whether that deviation is asymmetric ( ON or OFF ) or symmetric ( ON-OFF ) . However , the STC-NC is aimed primarily at high throughput cell classification , not spike prediction , where multiple-filter models ( including full STC analysis ) will provide greater accuracy [17] , [19] . For example , a single filter STC-NC LN model cannot describe the differing temporal dynamics sometimes observed in the separate ON and OFF RF filters of an ON-OFF cell ( Fig . 2G–H ) . But even with the limitation of a single filter to model ON- and OFF-type spikes , the STC-NC may form the foundation for more sophisticated , multi-filter or multi-step analyses . In addition to our standard 100 µm×100 µm checkers , we have also used 60 µm×60 µm checkers , which potentially provide better resolution in mapping the spatial RF ( Fig . S4 ) . The cells in Fig . S4A–B were exposed to both 60×60 µm and 100×100 µm checker stimuli for equal durations of time . As expected , the RF maps showed improved resolution with smaller checker sizes . However , as the checker size decreases , the response strength was also decreased with reduced spike counts and lower spike expectations ( Fig . S4A–B ) . Moreover , for a large number of cells , we were only able to map the RF with the 100×100 µm checkers because the smaller checkers did not elicit a strong enough response . For example , at P25 , we collected a total of 204 WT cells with RFs mapped by 100×100 µm checkers , but only 70 RFs were mapped with the 60×60 µm stimulus . Despite these limitations , we observed a similar development trend from P18 to P25 upon comparing the 60×60 µm and 100×100 µm checker data ( Fig . S4C ) . In WT , the RF sizes of ON and ON-OFF cells were decreased from P18 to P25 ( ON: p = 1 . 41×10−4; ON-OFF: p = 3 . 25×10−4 ) , but the size of OFF cells remains unchanged ( p = 0 . 59 in Wilcoxon rank sum test , Fig . S4C ) . We compared RF center measurements made by the STC-NC with those made by the STA ( e . g . [18] ) . Kerschensteiner and his collegues ( 2008 ) found that the RF 1σ radii for ON-center and OFF-center cells ranged from 60–120 µm in 2–3-month old mice using 66 µm×66 µm checkerboard stimuli . Here , we showed that at P18 , our data correspond to 1σ radii for ON-center and OFF-center RF centers of 70 µm and 67 µm , respectively ( Figs . 6–7 ) , which fall within the range of the published measurements [18] . Moreover , we demonstrate here that the STA provides inaccurate estimates of ON-OFF RF center sizes ( Figs 1–2 and Fig . S3 ) and that measurements based on the STC-NC are more accurate . The STC-NC better describes the ON-OFF character of the RF center , correlating well with the results of spot stimulation ( Fig . 3E ) , it reveals a well-structured spatial RF center ( Fig . 2F ) , and is capable of classifying cells efficiently into ON , OFF , and ON-OFF types ( Fig . 4A ) . The correlation between the RGC dendritic structure and its RF properties is not straightforward . Based on RGC dendritic morphology about 10–14 subtypes of RGCs have been identified in mouse [28]–[30] . Classification of RGCs based on physiological properties is incomplete , as is the correspondence between morphological and physiological types [1] , [2] . For some subtypes of RGCs a match between dendritic and RF properties can be made [31]–[34] . For example , one class of OFF RGCs has asymmetric dendritic arbors aligned in a dorsal-to-ventral direction across the mouse retina that matches their responses to visual stimuli moving in a soma-to-dendrite direction [32] . On the other hand , studies in the rabbit retina have shown that about 10% of the direction-selective cells have RFs displaced toward the preferred direction , while their dendritic structures exhibit no obvious corresponding relationship [35] . During postnatal development , RGC dendritic laminar refinement somewhat correlates with the functional separation of ON and OFF pathways . Using full-field flash stimulus , Tian and Copenhagen showed that ON-OFF cells decrease from 76% before eye-opening to 40% immediately after eye opening to 22% at P28 . In this study , we provide the reliable identification of the ON-OFF subtype and show that cells with ON-OFF centers decrease from 35% of the RGC population at P18 to 24% at P25 ( Fig . 6B ) . Based on RGC dendritic morphology , Landi et al . ( 2007 ) showed that 66% of RGCs were bistratified ( presumed ON-OFF ) at P10 , and this percentage decreased to 54% at P16 and 31% at P30 . These results are generally consistent with our previous studies of RGC dendritic laminar refinement where the percentage of RGCs possessing bi-laminated dendritic structure decreases from ∼50% to ∼35% from P13 to P28 [8] . In adult mice ( >P27 ) , about 50–60% RGCs are ON RGCs ( [8] , [11] , [30] , but see [28] which showed 50% of RGCs are ON-OFF RGCs ) . Unlike in cats , which have roughly 50–50 ON vs OFF cells [2] , much fewer OFF cells ( 5–15% ) are reported in mice [8] , [11] , [30] . In addition , RGC response can be also characterized along different dimensions other than ON vs . OFF , e . g . sustained vs . transient , and brisk vs . sluggish [2] . It is of great interest to characterize further these properties of RGCs by MEA with new analytical tools . At the same time , refinement of RGC dendritic arbor does not always correlate with the functional maturation . In the developing turtle retina , intense dendritic growth occurs before RGCs became sensitive to light , and a weak correlation is found between physiological RFs and dendritic arbor structure [36] . In kitten , the arbors of gamma RGCs are similar to their adult counterpart , while the dendritic fields of alpha cells in the peripheral retina reach their adult dimensions three weeks after birth , around which time beta cells begin to expand [6] , [7] . By contrast , most cells respond to light first at P10 with RF centers invariant or in some cases larger during postnatal development than in the adult [37] . In developing mouse retina , the size of the dendritic field is typically somewhat larger than the size of the RGC RF center determined by STC-NC analysis . For all RGCs , the dendritic field size at P13 ( mean: 17 . 6±1 . 1×103 µm2 ) is about 16% larger than the 1σ contour of their RF center at P18 ( mean: 14 . 8±0 . 3×103 µm2 ) . For ON RGCs , the mean dendritic field size at P13 is 16 . 2±1 . 3×103 µm2 [9] , about 12% larger than the mean 1σ area of their RF center at P18 ( 14 . 3±0 . 3×103 µm2 ) . Interestingly , during the two weeks after eye opening , the dendritic field size of ON RGCs increases almost 47% ( P28: 23 . 9±1 . 0×103 µm2 ) , while their 1σ RF center size at P25 ( 13 . 3±0 . 4×103 µm2 ) is 7% smaller than it was at P18 . These data suggest that the relationship between dendritic field extent and RF center size that seems to be quite robust in adult retina is less clear-cut during development when synaptic contacts are being established and refined . The developmental mechanisms of RGC dendritic structural refinement and their functional maturation remain to be elucidated . The early dendritic arborization and synaptic formation of RGCs are generally thought to be regulated by intrinsic growth programs [1] , [38] . For example , recent studies have shown that immunoglobulin superfamily ( IgSF ) adhesion molecules–Dscam , DscamL , Sidekick-1 and Sidekick-2–are expressed in distinct IPL sublaminae of chick retinae [39] . Loss- and gain-of-function studies in vivo showed that these IgSF members participate in determining the IPL sublaminae in which synaptic partners arborize and connect [39] . In later development , environmental signals are involved in the regulation of RGC dendritic maturation [1] , [8] , [38] . Time-lapse imaging experiments have revealed that RGCs take an active role in sampling the local retinal environment and in establishing functional synaptic contacts with amacrine and bipolar neurons by extending and retracting dendritic filopodia [40] . Selective removal of ON input causes a reduced rate of synapse formation rather than an increase in synapse elimination , creating ON-OFF RGCs with fewer synapses in their ON arbors without affecting OFF arbor structure [41] . Contrary to this view , other studies have suggested that the formation of synaptic connections between RGC dendrites and other neuronal processes in the IPL is established through the elimination of superfluous processes [42] , [43] . For example , large field type-I rat RGCs exhibit extensive branch loss [43] . Neurotrophins modulate dendritic development in many nervous systems [44] . We have previously shown that BDNF and NT-3 play overlapping roles in RGC dendritic laminar refinement and distinct roles in subtype-specific maturation of RGCs [8] , [9] . Indeed , many studies have suggested that BDNF can exert multiple roles on RGC dendritic development through different mechanisms [38] . For example , retinal BDNF levels directly affect the complexity of RGC dendritic arbors in Xenopus [8] , [45] . BDNF could also regulate the morphology and function of amacrine cells which in turn influence RGC dendritic connectivity indirectly [46] , [47] . Because the expression and release of BDNF is regulated by visual experience , BDNF may modify RGC dendritic laminar refinement by altering synaptic connectivity between RGCs and other input neurons [8] . Finally , target-derived BDNF could retrogradely affect RGC dendritic structure and function [45] , [48] . Compared to the intensive studies on BDNF , little is know about the roles of NT-3 in retinal development . In mouse retina , NT-3 is expressed both pre- and postnatally and its expression is unaffected by visual experience [9] , [49] . In slice cultures of cortical neurons , NT-3 stimulated dendritic growth in layer 6 and inhibited BDNF-stimulated dendritic growth in layer 4 [50] . In developing chick retina , NT-3 regulates RGC survival and the structure of the IPL [51] . Here we have shown that NT-3 regulates RF properties of RGCs during postnatal development . Our previous work showed that dendritic trees of RGCs that bistratify in the IPL are smaller in NT-3 OE mice , while those which are monostratified are not [9] . Here we have shown that the RF centers of ON-OFF but not ON or OFF center cells are smaller in adult NT-3 OE mice ( Fig . 8 ) , consistent with the earlier anatomical work . Future studies are needed to identify whether and how the refinement of dendritic field size and RF size is translated into the maturation of visual function in WT retinas and when the neurotrophin signaling is perturbed . In conclusion , we have established the STC-NC analysis as a method for the characterization of RGC subtype RF properties , and demonstrated that NT-3 regulates the functional development of ON-OFF center RGCs . Our study provides a basis for future examination of how neurotrophic signaling pathways modulate RGC RF properties .
Transgenic mice expressing NT-3 driven by the alpha A-crystallin promoter from the lens ( labeled as NT-3 OE mice ) on the BALB/c genetic background [52] were crossed more than ten times with C57BL/6 , so that the NT-3 OE mice were mainly on a C57BL/6 background [9] . All animal procedures conformed to the guidelines in the Use of Animals in Neuroscience Research from the NIH and were in accordance with protocols approved by the Northwestern University IACUC . NT-3 OE and WT mice were euthanized by direct cervical dislocation , and the eyes were placed into an oxygenated ( 95% O2 , 5% CO2 ) artificial cerebrospinal fluid ( ACSF ) . Under infrared illumination , a single retina was isolated and about one fifth of the retina was cut for each experiment . The retinal ganglion cell layer ( GCL ) was then placed into contact with a 60-channel multielectrode array ( Multichannel Systems Gmbh , Fig . S5 ) [13] . The retina was perfused with ACSF and maintained at 33–34°C during the entire recording session . An image from a computer monitor was projected onto the retina . As the retina responded to visual patterns on the monitor , voltage signals from the microelectrodes were amplified by the MCS preamplifier ( bandpass 1–5000 Hz ) and recorded ( MCRack software ) . Spike waveforms were collected using a voltage threshold and sorted with Plexon Offline Sorter to isolate spike trains [13] . Spike timestamps were then exported to Matlab , where custom analyses were applied individually to each RGC's spike train . To characterize the spatiotemporal visual responses of the RGCs at different developmental ages , three monochrome stimuli were applied: a full-field flash , a spatiotemporal Gaussian white noise , and a flashing spot stimulus . The mean luminance for all stimuli was 2 cd·m−2 . The full-field flash was a repetitive binary stimulus consisting of 2 seconds with the light OFF ( 0 cd·m−2 ) followed by 2 seconds with the light ON ( 4 cd·m−2 ) . The white noise stimulus appeared as a flickering grayscale checkerboard pattern with random spatial and temporal structure , which was composed of 100 µm×100 µm square checkers . A new random luminance was assigned to each checker every 33 ms . The Gaussian distribution from which luminance values were randomly drawn was centered at 2 cd·m−2 with a standard deviation of 0 . 78 cd·m−2 , causing the distribution to be truncated at ±∼2 . 5 standard deviations . The spot stimulus utilized the same checkerboard pattern as the white noise , however , the entire frame was assigned to mean luminance , and a single randomly chosen checker was flashed either ON ( 4 cd·m−2 ) or OFF ( 0 cd·m−2 ) every second . Both ON and OFF spots covered each location a minimum of 20 times , and the location and polarity of the flashed checker was randomized . Peristimulus time histograms ( PSTHs ) and raster plots of individual units were generated . The Response Dominance Index ( RDI ) was calculated from the transient peak spike rates during the first quarter of the ON ( RON ) and OFF portions ( ROFF ) of the full field stimulus by the following equation [11] , [22]: The value of the RDI ranges from −1 to 1 . Cells with an RDI near 1 possess an ON-dominating response , those with an RDI near −1 possess an OFF-dominating response , and the cells with an intermediate RDI near 0 possess a more balanced ON-OFF response . Spikes driven by the Gaussian white noise stimulus were placed into 33 ms bins defined by the stimulus refresh cycle . The 30 frames preceding each spike were linearized to form a vector , sn , which describes the spatiotemporal stimulus preceding the nth spike in the spike train . It has length , where f is the number of frames preceding the spike that are captured for analysis , and c is the number of checkers per frame . The Spike-Triggered Ensemble ( STE ) , which is the set of spatiotemporal visual stimuli preceding all spikes in a neuron's spike train , was constructed as an NxM matrix , S , by collecting all sn into the rows of S . The STE is a subset of the entire Raw Stimulus Set ( RSS ) , which is the collection of all the spatiotemporal stimuli , s , that were presented to the retina . Each element of the vector s possesses a value representing the deviation from mean luminance ( 2 cd·m−2 ) , and over all the s vectors in RSS , these values form a Gaussian distribution centered at zero . The STE and RSS collections exist in an M-dimensional space , which corresponds to the number of independent elements in s , and they often possess different means . The identification of statistical differences between the STE and the RSS in this M-dimensional space forms the basis for spike-triggered neural characterization [21] . The STA is simply the vector average of the STE . where N is the total number of spikes in the spike train [16] . Each element of the vector A , therefore , is the average of all the values stored in the corresponding column of S . The average of the RSS is the zero vector , but in general the STA differs greatly from zero , and it therefore represents a stimulus feature that evokes a spike from the neuron . During the analysis , the standard deviation of the elements in each cell's STA was calculated . A cell was considered to have a non-responsive STA if no single element of its STA exceeded six standard deviations in magnitude . In addition , we utilized the location of the STA receptive field to identify potential duplicate recordings of single RGCs on neighboring electrode channels . Crosscorrelation plots were generated to confirm and reject duplicate spike trains . In general , ∼40% duplicates for hexagonal MEA and for rectangular MEA ∼15% duplicates of spike trains with above-threshold ( “mappable” ) STAs were removed . STC analysis is implemented by performing a principal component analysis ( PCA ) on the STE . PCA can be achieved by eigendecomposition of the covariance matrix , which generates eigenvectors that are then sorted by their eigenvalues to identify directions of large and small variance . Because the STC estimation error is proportional to [21] , we windowed the stimulus vectors , sn , around the central checker of the STA . This step reduced the spatial extent of the stimulus , but did not affect the number of frames captured . We used a 5×5 window , and the effect was to create new sn vectors with smaller length , thereby reducing the dimensionality , M , of the stimulus . A shorter 666 ms ( 20 frames ) time period was also used . The covariance matrix , C , is an M×M matrix , and it was calculated with the following equation:C is a positive definite matrix as long as , otherwise it is positive semi-definite . Subsequently , eigendecomposition was performed on C , and the eigenvectors were sorted according to their corresponding eigenvalues . This process allowed us to find a new basis set for the stimulus space , in which the directions were ordered according to STE variance . Traditionally , during PCA , the covariance matrix is created by computing the outer product of the mean-centered STE with itself . Eigendecomposition is then performed on the covariance matrix , and the resulting eigenvectors represent the principal components . However , in situations where classification is the primary goal , eigendecomposition of a non-centered moment matrix can be justified when the zero vector is an important point of reference [24] , [27] . We computed the non-centered second moment matrix M , with the following equation: In the case of a non-centered moment matrix , the eigenvector with the greatest eigenvalue maximizes the second moment of the STE around zero – not the variance , which is the second moment around the mean . The resulting eigenvectors are often referred to as non-centered principal components [27] . It is important to note that since the covariance matrix is not used , these eigenvectors are not strictly principal components in the STC-NC technique . Because the STC-NC is only a 5×5 window , the RF makes up a large portion of the STC-NC vector . For this reason , to get a better measure of how strong the RF center signal was compared to the surrounding noise , the maximum contrast of the STC-NC vector was divided by the standard deviation of the outermost ring of pixels in the STC-NC ( not the standard deviation of the full vector ) . In the analysis , we excluded RGCs with low STC-NC signal which had visual responses that were weak or non-existent . The code for the core STC-NC algorithm , along with sample data , is available at: http://code . google . com/p/non-centered-spike-triggered-covariance/ In a LN model for RGC stimulus-response transduction , the scalar output from one or several linear filters is used as input for a static nonlinearity , which predicts the probability that the neuron will spike . The static nonlinearity is a function that maps from k-dimensional space to a 1-dimensional space ( ℜk ⇒ ℜ1 ) , where k is the number of linear filters used . We primarily used a single filter in this study . Thus , the nonlinear transform was computed in a few simple steps [16] , [21] . First the output of the linear filter for all s in RSS was calculated and binned to form a vector LRSS . For Gaussian white noise , the resulting distribution very closely approximated the Gaussian distribution used for stimulus generation . Next , the output of the linear filter for all sn in the STE was calculated and binned to form LSTE . Finally , LSTE was divided , element-by-element , by LRSS . The resulting vector , N , held the fractional number of spikes expected in response to a given input from the linear filter . The vector N deviates from predicting spike probability only in the sense that the values in this vector can exceed 1 if more than 1 spike is expected within the binned time period . Prior to this calculation , we multiplied all OFF-type linear filters by −1 , forcing them to appear as ON-type filters . This trick ensures that all stimuli that represent a light offset in the RF center create a negative scalar output when convolved with the linear filter , and vice versa for stimuli representing light onset in the RF center . This was done to properly orient the polarity of the 1D STE and RSS projections onto the linear filters , which in turn orients the static nonlinearities in a standardized way independent of the linear filter character so that ON , OFF , and ON-OFF cells can be easily identified and compared . For best results in this step only , the STA , STC , and STC-NC linear filters , as well as the stimulus vectors sn , were further reduced using a 300 µm×300 µm checker window . Because computing LRSS is extremely computationally intensive , and because LRSS did not vary significantly for each linear filter , we used an LRSS that was averaged over 10 randomly chosen linear filters for the computation of all N vectors . Because the Gaussian distribution for our white noise was truncated at ∼2 . 5 standard deviations , the values of the N vector corresponding to LRSS and LSTE inputs beyond 2 . 5 standard deviations of the LRSS distribution were discarded . The ON/OFF/ON-OFF bias of the cell was then determined with the scalar value:where PON and POFF represent integrals of the static nonlinearity over the positive or negative ranges , respectively . We labeled a cell OFF-center with , ON-OFF center with , and ON-center with . RF center areas were computed from the STA and the STC-NC using two methods . In the first method , the single frame possessing the maximal deviation from the mean was isolated , and a bivariate Gaussian distribution was fit to this frame . The bivariate Gaussian function was described by:where A is the amplitude of the Gaussian , x and y are independent spatial coordinates , h and k are the x- and y-coordinate receptive field midpoints respectively , σa and σb are the standard deviations of the major and minor axes respectively , and θ is the angle between the major axis of the Gaussian and the x-axes of the coordinate system [53] . From the least squares fit of this distribution to the linear filter frame possessing maximal deviation from mean luminance , the RF center area within the 1σ ellipse of the Gaussian distribution was calculated according to . Because the STC-NC was windowed prior to computation , in order to improve fitting results , we loosely forced the Gaussian to zero at the edges of the STC-NC by framing it in a large surrounding zero-contrast border . A 6σ signal strength threshold was used to eliminate weakly responsive cells from further analysis . The second method directly counted checkers in the frame of maximal deviation with amplitudes exceeding 0 . 35× the maximal magnitude . Although these two methods are not completely analogous , a comparison of the results shows that they are strongly correlated ( p<0 . 001 , R = 0 . 83; Fig . S6 ) . In a subset of the experiments , spot stimuli were applied in conjunction with Gaussian white noise . STA RF mapping was performed to identify the checker that most closely approximated the RF center . PSTHs were then generated to determine the cell's response to ON and OFF flashes in the single central checker . The peak firing rates in the first 500 ms following the ON and OFF flashes were used to calculate the Spot Bias with an equation analogous to that used in the calculation of the RDI:but now RON is the peak firing rate following the onset of the light spot , and ROFF is the peak firing rate following the onset of the dark spot . Comparison of distribution means was performed using the Student's t-test . Because the shape of the RF center size distributions departed significantly from Gaussian , a Wilcoxon rank sum test was used to compare RF center size medians . We also utilized the Analysis of Covariance technique ( ANCOVA ) to remove the effect of the confounding STC-NC signal strength prior to calculating the significance of differences in RF size due to grouping . The Kolmogorov-Smirnov ( KS ) test was used to compare the shapes of distributions of continuous-valued variables ( such as STC-NC Bias ) , and a two sample χ2 test was used to compare distributions of categorical variables ( such as ON , OFF , and ON-OFF classifications ) . Correlations were described using Pearson's linear correlation coefficient , and corresponding p-values were calculated using the Fisher transformation to map the correlation coefficient onto a t-statistic . Box plots utilized Matlab defaults , in which notch widths representing the 95% confidence interval for the median were calculated [54] , and outliers were defined as data points lying 1 . 5 times the interquartile range beyond the upper or lower quartiles . To test for bimodality in the spike-triggered ensemble , the Hartigan and Hartigan Dip test was used , and p-values against the null hypothesis of unimodality were calculated by bootstrapping samples of the appropriate size drawn from a uniform distribution [55]–[58] . Retinas were dissected and fixed in 4% ( w/v ) paraformaldehyde in 0 . 1 M phosphate-buffered saline ( pH 7 . 4 ) at P16 [8] . Cryostat sections or whole-mounted retinas were prepared as described in Liu et al . ( 2007 ) . The primary antibodies include anti-mouse Brn-3a ( 1∶100 , Chemicon International ) and anti-mouse SMI-32 ( 1∶1000 , Sternberger Monoclonal Inc . ) . For confocal microscopy , images were captured with a Zeiss Pascal confocal microscope ( Zeiss , Thornwood , NY ) [8] . For cell counting , immuno-positive cells from 6–10 fields from each retina were counted and the average density was calculated in LSM5 Image browser ( Zeiss ) or ImageJ . | The developmental separation of ON and OFF pathways is one of the fundamental features of the visual system . In the mouse retina , some bi-stratified ON-OFF RGCs are refined into mono-stratified ON or OFF RGCs during the first postnatal month . However , the process by which the RGCs' physiological receptive field properties mature remains incompletely characterized , mainly due to the lack of a reliable and efficient method to classify RGCs into different subtypes . Here we have developed an innovative analysis , Spike Triggered Covariance – Non-Centered ( STC-NC ) , and demonstrated that this technique can accurately characterize the receptive field properties of ON , OFF and ON-OFF center cells . We show that , in wildtype mouse , RGCs gradually develop sensitivity to focal stimulation after eye opening , and the development of ON-OFF receptive field center properties correlates well with their dendritic laminar refinement . Furthermore , overexpression of NT-3 accelerates the developmental decrease of receptive field center size in ON-OFF cells . Our study is the first to establish the STC-NC analysis which can successfully identify ON-OFF subtype RGCs and to demonstrate how receptive field development relates to a neurotrophic driver in the retina . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"physiology/sensory",
"systems",
"physiology/neuronal",
"signaling",
"mechanisms",
"computational",
"biology/computational",
"neuroscience",
"neuroscience/neurodevelopment",
"developmental",
"biology/neurodevelopment"
] | 2010 | Non-Centered Spike-Triggered Covariance Analysis Reveals Neurotrophin-3 as a Developmental Regulator of Receptive Field Properties of ON-OFF Retinal Ganglion Cells |
RNA silencing is a highly conserved pathway in the network of interconnected defense responses that are activated during viral infection . As a counter-defense , many plant viruses encode proteins that block silencing , often also interfering with endogenous small RNA pathways . However , the mechanism of action of viral suppressors is not well understood and the role of host factors in the process is just beginning to emerge . Here we report that the ethylene-inducible transcription factor RAV2 is required for suppression of RNA silencing by two unrelated plant viral proteins , potyvirus HC-Pro and carmovirus P38 . Using a hairpin transgene silencing system , we find that both viral suppressors require RAV2 to block the activity of primary siRNAs , whereas suppression of transitive silencing is RAV2-independent . RAV2 is also required for many HC-Pro-mediated morphological anomalies in transgenic plants , but not for the associated defects in the microRNA pathway . Whole genome tiling microarray experiments demonstrate that expression of genes known to be required for silencing is unchanged in HC-Pro plants , whereas a striking number of genes involved in other biotic and abiotic stress responses are induced , many in a RAV2-dependent manner . Among the genes that require RAV2 for induction by HC-Pro are FRY1 and CML38 , genes implicated as endogenous suppressors of silencing . These findings raise the intriguing possibility that HC-Pro-suppression of silencing is not caused by decreased expression of genes that are required for silencing , but instead , by induction of stress and defense responses , some components of which interfere with antiviral silencing . Furthermore , the observation that two unrelated viral suppressors require the activity of the same factor to block silencing suggests that RAV2 represents a control point that can be readily subverted by viruses to block antiviral silencing .
Plants have a complex interconnected system of defense and stress pathways [1] , [2] that receives incoming stimuli , transduces the signal and initiates the appropriate response . The process is orchestrated by a variety of plant hormones and small signaling molecules , and the final shape of the response is refined by crosstalk among different pathways in the network . Evidence emerging over the last decade has made it clear that RNA silencing and endogenous small RNA pathways constitute a major response to a variety of biotic and abiotic stresses [3] , [4] , [5] . Surprisingly , however , although many of the components of the silencing machinery are known , little is yet known about how silencing is regulated or how it is integrated into the network of other defense and stress pathways . RNA silencing is a sequence specific RNA degradation mechanism that serves an important antiviral role in plants [6] . Antiviral silencing is triggered by double stranded RNA ( dsRNA ) that arises during virus infection . The dsRNA trigger is processed by DICER-LIKE ( DCL ) ribonucleases into primary short interfering RNAs ( siRNAs ) , which incorporate into an ARGONAUTE ( AGO ) protein-containing effector complex and guide it to complementary target RNAs . The destruction of target RNAs can be amplified via a process called transitive silencing , in which the target RNA serves as template for host RNA-dependent RNA polymerases ( RDRs ) to produce additional dsRNA that is subsequently processed into secondary siRNAs . In addition to these RDRs , a number of other genes , including DCL2 , AGO1 and SUPPRESSOR OF GENE SILENCING 3 ( SGS3 ) , are required for transitive silencing , but not for primary silencing [7] , [8] , [9] . The primary and transitive silencing pathways work together to limit the accumulation of viral RNAs during both the initial and systemic phases of infection . In addition to antiviral silencing and related pathways that target invading nucleic acids , there are endogenous small RNA pathways that regulate gene expression by directing cleavage of target RNA , inhibition of mRNA translation , or modification of chromatin structure . The best studied of the endogenous small RNAs are the microRNAs ( miRNAs ) , which play major roles in development and in response to a variety of stresses [10] , [11] , [12] . Although different small RNA mediated pathways have unique genetic requirements , all make use of an overlapping set of genes for their biogenesis ( four DCL genes ) and function ( ten AGO genes ) , and there is growing evidence that these pathways are interconnected and compete with one another . For example , DCL1 , the Dicer that produces most miRNAs , represses antiviral silencing by down-regulation of DCL3 and DCL4 [13] and , when over-expressed , blocks silencing induced by a sense transgene [14] . In addition , many viral suppressors of RNA silencing also interfere with the biogenesis and/or function of endogenous small RNAs such as miRNAs and trans-acting small interfering RNAs ( tasiRNAs ) [15] , [16] . However , the mechanisms that regulate and integrate the various small RNA pathways are just beginning to be elucidated . Plant viruses have evolved a variety of effective counter-defensive strategies to suppress silencing . Numerous plant viruses encode proteins that block some aspect of RNA silencing [15] , [16] . These viral proteins are highly diverse in primary sequence and protein structure , though they may share certain mechanistic features . For example , the ability to bind small RNAs is a feature of many viral suppressors of silencing , including the two used in the present work . Indeed , it has been proposed that most viral suppressors of silencing work by binding and sequestering small RNAs , thereby blocking their activity [17] , [18] . However , the physiological significance of small RNA binding is not yet clear in many cases [6] , and some suppressors manipulate silencing via interaction with host proteins that are either components of the silencing machinery [19] , [20] , [21] , [22] or proposed regulators of the pathway [23] . Thus , the mechanism of action of viral suppressors is likely both diverse and complex and is not yet fully understood . Our studies have focused on understanding the mechanism of action of HC-Pro , a potent viral suppressor of silencing that blocks both primary and transitive silencing . Our approach has been to identify host proteins that physically interact with HC-Pro and examine the effect of altering the levels of these proteins on both RNA silencing and the ability of HC-Pro to block silencing [23] . Using this approach , we find that RAV2/EDF2 ( hereafter referred to as RAV2 ) , an HC-Pro-interacting protein that is a member of the RAV/EDF family of transcription factors , is required for suppression of silencing not only by potyvirus HC-Pro , but also by carmovirus P38 , the silencing suppressor from a virus family unrelated to potyviruses . Interestingly , RAV2 is required exclusively for blocking the activity of primary siRNAs , whereas suppression of transitive silencing and effects on the endogenous microRNA pathway are RAV2-independent . Whole genome tiling microarray experiments were used to characterize HC-Pro-mediated changes in host expression and identify which , if any , were RAV2-dependent . The results raise the interesting possibility that HC-Pro-suppression of silencing is not caused by decreased expression of genes that are required for silencing , but instead , by induction of stress and defense pathways that interfere with antiviral silencing .
In previous work we used a yeast two-hybrid screen to identify Nicotiana tabacum proteins that interact with Tobacco Etch Potyvirus ( TEV ) HC-Pro [23] . One of the proteins identified in this way was named ntRAV because of its relatedness to the Arabidopsis thaliana RAV/EDF family of transcription factors . The RAV/EDF protein family has six members , and these are unique among transcription factors in having two unrelated DNA binding domains ( AP2 and B3 ) [24] . Members of this family are responsive to numerous biotic and abiotic stresses [25] , [26] , [27] , [28] and are inducible by the plant hormone ethylene [29] , which controls many aspects of plant physiology , including defense against pathogens [30] , [31] . In vitro pull-down experiments were used to confirm a physical interaction between TEV HC-Pro and ntRAV . 35S-methionine-labeled ntRAV produced in a coupled in vitro transcription/translation system co-purified with an HC-Pro-GST fusion protein isolated from recombinant bacteria , but not with GST alone ( Fig . 1A and see also Fig . S1 and Text S1 ) . This result validates the HC-Pro-ntRAV interaction initially identified in the yeast two-hybrid system . To determine if ntRAV plays a role in RNA silencing , we evaluated the effect of ntRAV over-expression on transgene-induced silencing . In tobacco , ntRAV is normally expressed at high levels throughout fully expanded healthy leaves of young plants , but expression decreases greatly starting at about 24 days after germination ( Fig . 1B , lanes 1–6 ) . In contrast , a tobacco line that ectopically expresses ntRAV from the constitutive Cauliflower mosaic virus ( CaMV ) 35S promoter maintains high level expression of ntRAV ( Fig . 1B , lanes 7–9 ) . We crossed the 35S:ntRAV transgenic line , as well as wild type and HC-Pro-expressing control lines , to the well-characterized tobacco transgenic line 6b5 [32] , which is post-transcriptionally silenced for a transgene encoding β-glucuronidase ( GUS ) . Silencing of the GUS locus in line 6b5 reinitiates every generation , starting in the vascular tissue of the oldest leaves and then spreading throughout the leaf . The expression of GUS in F1 progeny of these crosses was assayed histochemically in leaves ( Fig . 1C ) and by northern blots of RNA from the vascular tissue ( Fig . 1D ) at 26 days after germination . In these young plants , ectopic expression of ntRAV blocked silencing of GUS in vascular tissue of fully expanded , healthy leaves about as well as HC-Pro ( Fig . 1C and D ) . However , unlike HC-Pro , which completely blocks silencing over the lifetime of the plant , ectopic expression of ntRAV only delayed the onset of silencing , and GUS was eventually silenced throughout the leaf ( data not shown ) . These results , together with those showing a physical interaction between ntRAV and TEV HC-Pro proteins , raised the possibility that ntRAV plays a role in HC-Pro-mediated suppression of silencing . To further investigate the role of ntRAV in HC-Pro suppression of silencing , we switched from tobacco to Arabidopsis thaliana , in order to take advantage of the numerous genetic tools available in that model system . Our experiments focused on a RAV gene family member closely related to the tobacco ntRAV , Arabidopsis RAV2 ( At1g68840 ) , which had already been cloned and characterized , and for which a validated T-DNA insertional knockout line was available [29] . The change in experimental system also necessitated a change from the HC-Pro encoded by TEV to that encoded by turnip mosaic virus ( TuMV ) , a related potyvirus that infects Arabidopsis . Like the TEV HC-Pro transgene in tobacco , expression of the TuMV HC-Pro in transgenic Arabidopsis plants has been shown to suppress both virus- and transgene-induced RNA silencing [14] , [33] . The TuMV HC-Pro transgenic line used in our experiments expresses HC-Pro at a high level and is highly phenotypic [14] . We used in vivo pull-down experiments to determine whether the TuMV HC-Pro and RAV2 proteins interact , as would be expected if RAV2 were a functional homolog of ntRAV . In these experiments , the homozygous rav2 knockout line [29] was transformed with a construct designed to express a transgene encoding FLAG-tagged RAV2 . A transformant that expressed the FLAG-RAV2 transgene was crossed to our TuMV HC-Pro transgenic line [14] , and expression of both transgenes in the F1 offspring was confirmed by RNA gel blot analysis ( data not shown ) . Pull-down experiments using antiserum specific to the FLAG tag , followed by western blot analysis , showed that TuMV HC-Pro co-immunoprecipitates with the Flag-tagged RAV2 ( Fig . 2 ) , indicating that RAV2 and TuMV HC-Pro interact in planta in Arabidopsis . This result confirms that RAV2 is a functional homolog of ntRAV and also provides evidence that the interaction between potyviral HC-Pro and host RAV-like transcription factors is a conserved feature of these proteins . Our initial experiments to examine the role of RAV2 in HC-Pro suppression of silencing focused on VIGS . These experiments used the well characterized geminivirus silencing vector , cabbage leaf curl virus ( CaLCV ) , which carried a portion of the endogenous CHLORATA42 ( CH42 ) gene [34] . CH42 is required for chlorophyll accumulation , and VIGS of CH42 in wild type plants results in extensive chlorosis and marked reduction in the level of CH42 mRNA . These changes are accompanied by a pronounced accumulation of 24-nt siRNAs that derive from the CH42 sequences within the viral vector [14] , [34] . HC-Pro transgenic plants become infected when bombarded with the CH42 VIGS vector and , although high levels of siRNAs accumulate in the plants , the CH42 gene is not silenced as evidenced by accumulation of CH42 mRNA and the absence of chlorosis [14] . To determine if RAV2 is required for HC-Pro suppression of VIGS , plants expressing HC-Pro in either the wild type or the rav2 knockout background , along with control plants , were bombarded with the CH42 VIGS vector . Wild type control plants as well as rav2 knockout plants exhibited chlorosis of infected tissues ( Fig . 3A , top two panels ) accompanied by reduction in CH42 mRNA levels and the concomitant accumulation of siRNAs , as expected for VIGS ( Fig . 3B , lanes 1–4 ) . HC-Pro transgenic plants were suppressed for VIGS of CH42 , remaining green ( Fig . 3A , bottom left panel ) and accumulating wild type levels of CH42 mRNA as previously reported ( Fig . 3B , lanes 5 and 6 ) . In contrast , HC-Pro transgenic plants in the rav2 knockout background were competent for VIGS of CH42 as evidenced by systemic chlorosis ( Fig . 3A , bottom right panel ) accompanied by reduction in CH42 mRNA levels ( Fig . 3B , lanes 7–9 ) . This result indicates that RAV2 is required for HC-Pro suppression of VIGS . To examine the role of RAV2 in HC-Pro-suppression of transgene silencing , we used a well-characterized system in which silencing occurs through both the primary and transitive branches of the silencing pathway [7] , [35] . This system is composed of two transgenes , the 306 and 6b4 loci ( Fig . 4A ) . The 6b4 locus encodes an expressing GUS transgene that includes the entire GUS coding sequence , while the 306 locus encodes a hairpin construct designed to silence GUS expression . The GUS sequence in the 306 locus has a 231 nucleotide deletion in the coding region ( Fig . 4A , shown in green ) so that RNAs originating from the 6b4 transcript can be unambiguously distinguished . The primary and transitive branches of silencing can be easily differentiated in this system . Basically , primary siRNAs derive only from the stem of the 306 hairpin transcript ( Fig . 4A , shown in red , probe 1 ) , whereas secondary siRNAs arise from either locus during an RDR6-dependent process called transitive silencing . In the case of the 306 transgene , siRNAs that arise from the loop of hairpin transcript are secondary siRNAs ( Fig . 4A , shown in blue , probe 3 ) . In contrast to the 306 hairpin transcript , the 6b4 mRNA produces only RDR6-dependent secondary siRNAs ( Fig . 4A , shown in red , green and blue; [7] . Thus , in the 306/6b4 system , 6b4 mRNA can be degraded by two mechanisms . It can be targeted by a RISC complex directed by siRNAs , or it can be a substrate for RDR6 , producing dsRNA that is subsequently processed by DCL to produce secondary siRNAs via transitive silencing . HC-Pro suppresses silencing in the 306/6b4 system , but has different effects on primary and secondary siRNAs: accumulation of secondary siRNAs is eliminated , as shown by the failure to detect any siRNAs when using either probe 2 or probe 3 [7] . In contrast , high levels of primary siRNAs accumulate , but are unable to mediate degradation of the 6b4 target RNA [7] . To determine if RAV2 is required for HC-Pro suppression of hairpin transgene silencing , we crossed the homozygous rav2 knockout line to a transgenic line homozygous for the 306 and 6b4 loci and hemizygous for the TuMV HC-Pro locus . F1 offspring of this cross were allowed to self-fertilize , producing an F2 population that was segregating for all four loci . F2 plants were genotyped , and individuals containing the 306/6b4/HC-Pro loci in the homozygous rav2 mutant background were identified , along with control plants containing all three loci in the wild type RAV2 background . The absence of RAV2 mRNA in rav2 knockout plants was verified by RNA gel blot analysis ( Fig . 4B ) . Initial analysis of the 306/6b4/HC-Pro plant lines addressed the possibility of transcriptional gene silencing ( TGS ) of the three transgenes involved , all of which are under the control of the CaMV 35S promoter . This was especially important because it has been shown that T-DNA insertion mutants that carry 35S promoter sequences , such as the rav2 knockout line used in this work , can induce TGS of other 35S promoters in the genome [36] and because HC-Pro cannot suppress silencing at the transcriptional level [37] , [38] . RNA gel blot analysis showed that the level of HC-Pro mRNA was similar in all plants carrying the HC-Pro transgene ( Fig . 4B ) , arguing against transcriptional silencing of 35S promoter sequences in the plants . In addition , the presence of siRNAs that derive from the GUS transcripts ( Fig . 4C ) indicates that the observed silencing of the GUS transgenes is at the post-transcriptional rather than the transcriptional level . The role of RAV2 in HC-Pro suppression of hairpin transgene silencing was assayed using northern blot analysis to measure the accumulation of 6b4 GUS target mRNA as well as that of GUS primary and secondary siRNAs ( Fig . 4C ) . As previously reported [7] , HC-Pro blocked target RNA degradation when 306/6b4/HC-Pro transgenic plants were wild type for RAV2 , showing the characteristic absence of secondary siRNAs accompanied by high levels of nonfunctional primary siRNAs ( Fig . 4C , compare lanes 3 and 4 ) . In contrast , HC-Pro failed to prevent degradation of the 6b4 GUS mRNA target in the rav2 knockout background ( Fig . 4C , lanes 1 and 2 ) . In addition , accumulation of GUS primary siRNAs was reduced in the rav2 compared to the RAV2 background and was similar to that in 306/6b4 plants without HC-Pro ( Fig . 4C , lanes 1–4 ) . Accumulation of secondary siRNAs , which are diagnostic of transitive silencing , was suppressed in HC-Pro transgenic plants even in the rav2 knockout background ( Fig . 4C , lanes 1–3 ) , suggesting that HC-Pro-suppression of transitive silencing is RAV2-independent . In this experiment , however , we cannot rule out the possibility that the rav2 knockout itself eliminates accumulation of secondary siRNAs . Therefore , our results suggest that RAV2 is required for the HC-Pro-mediated block in primary siRNA activity , but not for HC-Pro suppression of transitive silencing . To determine if RAV2 plays a general role in viral suppression of silencing , we used the 306/6b4 hairpin transgene silencing system to investigate whether Turnip Crinkle Virus ( TCV ) P38 , a viral suppressor of silencing from a different virus family than TuMV HC-Pro [39] , requires RAV2 to block silencing . The rav2 knockout line was crossed to a 306/6b4 line that expresses P38 , and the resultant F1 plants were allowed to self-fertilize . F2 plants were genotyped , and individuals containing the 306/6b4/P38 loci in the homozygous rav2 mutant background were identified along with control plants containing all three loci in the RAV2 background . We used northern blot analysis to confirm the expected pattern of expression of RAV2 and P38 in these two sets of plants ( Fig . 4D ) and to examine suppression of silencing by P38 in the presence and absence of RAV2 . Previous experiments showed that P38 behaves much like HC-Pro in the 306/6b4 transgene silencing system , blocking silencing and allowing 6b4 GUS mRNA to accumulate , even though high levels of GUS primary siRNAs also accumulate [7] . Similar to HC-Pro , P38 also blocks transitive silencing in this system as indicated by the absence of GUS secondary siRNAs [7] . In the current work , P38 transgenic 306/6b4 plants with at least one copy of the wild type RAV2 locus replicated those earlier results , showing P38 suppression of silencing , with a concomitant increase in accumulation of GUS primary siRNAs and elimination of GUS secondary siRNAs ( Fig . 4E , compare lanes 3 and 4 ) . In contrast , P38 suppression of silencing was strongly diminished in the rav2 knockout background ( Fig . 4E , lanes 1 and 2 ) . Similar to our results with HC-Pro , accumulation of primary siRNAs in plants expressing P38 was much reduced in the rav2 compared to the RAV2 background , whereas secondary siRNA accumulation was unaffected by the loss of RAV2 and remained undetectable ( Fig . 4E , compare lanes 1 and 2 with lane 3 ) . The variability in accumulation of primary siRNAs observed in the rav2 background ( Fig . 4E , lanes 1 and 2 ) probably reflects the facts that individual plants were tested and accumulation of primary siRNAs is greatly reduced , but not eliminated in the absence of RAV2 . Altogether our results indicate that RAV2 plays similar roles in suppression of silencing by P38 and HC-Pro . Interestingly , in both cases , RAV2 function is required for suppression of primary siRNA-directed target degradation , but dispensable for the block to transitive silencing . Arabidopsis plants expressing TuMV HC-Pro display a number of developmental anomalies: the plants are dwarfed with serrated leaves and have abnormal flower morphology associated with severely reduced fertility ( Fig . 5A; [14] , [33] ) . The phenotype of homozygous rav2 knockout plants , however , is indistinguishable from that of wild type plants ( data not shown ) . To determine if RAV2 is required for any of the HC-Pro associated developmental anomalies , we compared the phenotype of HC-Pro plants in the wild type RAV2 background to that of plants expressing approximately equal levels of HC-Pro mRNA , but in the rav2 knockout background . The HC-Pro-mediated defects in flower morphology and fertility are completely alleviated in the absence of RAV2 ( Fig . 5 and data not shown ) . In addition , both the dwarfing and serrated leaf phenotypes are mitigated - but not eliminated - in the rav2 knockout background , resulting in an intermediate phenotype that is most visible when the plants are young ( Fig . 5A ) , but becomes less distinguishable from that of wild type after the plants have flowered ( Fig . 3A , 5A , and data not shown ) . These observations indicate that RAV2 is required for HC-Pro-mediated flower and fertility defects and contributes to the defects in plant size and leaf shape . In addition to its role in suppression of silencing , HC-Pro also causes defects in the biogenesis and function of certain endogenous small RNAs , including miRNAs , a class of small regulatory RNAs that plays critical roles in development . MiRNAs arise by processing of stem-loop primary transcripts by a Dicer-like enzyme , usually DCL1 . The initial product is a 21-nt duplex , composed of the mature miRNA and the imperfectly complementary opposite strand , which is called miRNA* . The two strands separate and the mature miRNA binds to an AGO protein , forming the core of the miRNA effector complex . In HC-Pro transgenic plants , the level of many miRNAs is increased , often dramatically [33] , [40] . Despite the increased level of the miRNA in the HC-Pro plants , the miRNA-targeted messenger RNAs also show an increased accumulation , suggesting that the miRNAs have reduced function [33] , [41] . In addition , the miRNA* strand , which is unstable and fails to accumulate in wild type plants , characteristically accumulates to high levels in HC-Pro transgenic plants [33] . Together these results have led to the idea that HC-Pro impedes the proper separation of the strands of the miRNA:miRNA* duplex , leading to reduced association of the mature miRNA with AGO and thereby reducing miRNA function . Because RAV2 is required for HC-Pro effects on the biogenesis and function of primary siRNAs , as well as for many of the HC-Pro-associated developmental anomalies , we hypothesized that RAV2 might also be required for HC-Pro-mediated defects in the miRNA pathway . To address the role of RAV2 in HC-Pro-associated defects in miRNA biogenesis , we compared the levels of a variety of miRNAs and their corresponding miRNA* strands in HC-Pro plants in the presence and absence of RAV2 . In all cases , the levels of miRNA and miRNA* were independent of RAV2 ( Fig . 5B ) . These results indicate that RAV2 is not required for the HC-Pro-associated defects in miRNA biogenesis . To determine if RAV2 is involved in HC-Pro-associated defects in miRNA function , we compared the levels of a set of known miRNA-targeted messenger RNAs in RAV2/HC-Pro plants to those in rav2/HC-Pro plants using whole genome tiling microarray data ( see following section for details of the tiling array experiments ) . Because HC-Pro interferes with the activity of some miRNAs [33] , [41] , we expected the tiling array data to show increased expression of at least some miRNA-targeted genes in HC-Pro plants . The tiling array data supported this expectation . Specifically , out of 146 verified miRNA targets [42] , [43] , [44] , [45] , we found that 39 showed altered expression in the HC-Pro transgenic line compared to the wild type control . Of these , 35 had increased expression , and only one of these was up-regulated in HC-Pro/RAV2 versus HC-Pro/rav2 plants ( Table S1 ) , suggesting that RAV2 does not play a general role in HC-Pro inhibition of miRNA activity . Altogether , the results suggest that , although RAV2 is required for many of the morphological anomalies in HC-Pro transgenic plants , it is not required for the HC-Pro-mediated defects in either the biogenesis or function of miRNAs . Because RAV2 is a transcription factor , we expected that it might be required for some HC-Pro-mediated changes in gene expression and that identifying these genes could provide insight into the role of RAV2 in HC-Pro suppression of silencing . To address this idea , we employed whole genome tiling microarray experiments to determine if the global pattern of gene expression is altered in HC-Pro transgenic plants and , if so , whether any of the changes are dependent on RAV2 function . Arabidopsis plants with four different genotypes were used in this experiment: 1 ) a rav2 mutant line , 2 ) an HC-Pro expressing line , 3 ) the rav2 mutant line expressing HC-Pro , and 4 ) the wild type ( Columbia ecotype ) control . We grew all four genotypes under identical conditions , extracted total RNA from plants just before bolting and used poly-A RNA to generate probes for hybridization to the Arabidopsis tiling arrays as previously described [46] , [47] . TileMap [48] was used to identify genes that are significantly up- or down-regulated in each line as compared to wild type plants , as well as to compare the pattern of gene expression in RAV2/HC-Pro plants versus rav2/HC-Pro plants ( Tables S2–S9 ) . To check the tiling results , the expression of ten genes in these plant lines was additionally examined using real-time quantitative PCR ( RT qPCR ) . This analysis confirmed the relative levels of expression of these genes determined by the tiling array in 33 of 40 two-way comparisons between the four genotypes ( Fig . 6A and B and Fig . S2 ) . One of the first questions we addressed was whether genes involved in antiviral silencing and other small RNA pathways were affected by HC-Pro and RAV2 . Unexpectedly , none of the genes encoding components of the silencing machinery or otherwise known to be required for silencing were down-regulated in the HC-Pro plants . Expression of RAV2 itself was also not altered in HC-Pro plants . However , a number of silencing-associated genes were up-regulated in HC-Pro plants . The up-regulated genes included three of the ten Arabidopsis AGO family members , AGO2 , AGO3 , and AGO4 . AGO4 is required for some kinds of transcriptional silencing . The roles of AGO2 and AGO3 are unknown , but neither has been associated with antiviral silencing [49] , [50] . Interestingly , two genes implicated as endogenous suppressors of silencing were also up-regulated in HC-Pro: Arabidopsis FIERY1 ( FRY1 ) , which negatively regulates transitive silencing [51] , and CML38 ( At1g76650 ) , which is a likely Arabidopsis homolog of rgsCaM , an endogenous suppressor of antiviral silencing in tobacco [23] . Like RAV2 , rgsCaM was originally identified as an HC-Pro interacting protein [23]; however , it is not yet known whether rgsCaM is required for HC-Pro to suppress silencing . RT qPCR confirmed the relative expression levels of AGO2 , FRY1 , and CML38 in the HC-Pro expressing line compared to wild type plants ( compare Fig . 6A and 6B ) . The RT qPCR data also showed that increases in both FRY1 and CML38 expression required RAV2 , whereas the increase in AGO2 expression was only partially dependent on RAV2 ( Fig . 6A ) . These results argue that the mechanism for HC-Pro suppression of silencing does not involve down-regulation of genes required for silencing , but rather a RAV2-dependent up-regulation of genes that potentially antagonize antiviral silencing . The tiling array analysis was used to identify global HC-Pro-mediated changes in gene expression and determine which , if any , depended on RAV2 . A significant number of genes were differentially regulated in the HC-Pro plants; 2580 were up-regulated ( Table S2 ) and 2060 were down-regulated ( Table S3 ) . Many fewer genes were differentially affected in RAV2/HC-Pro compared to rav2/HC-Pro plants ( Tables S4 and S5 ) . Of 265 genes that showed dependence on RAV2 for up-regulation by HC-Pro ( Table S10 ) , only a small number showed changed expression in rav2 mutant plants in the absence of HC-Pro as compared to wild type ( 20 of 265 were up-regulated; 17 of 265 were down-regulated ) . Similarly , of 433 genes that showed dependence on RAV2 for down-regulation by HC-Pro ( Table S11 ) , a relatively small number showed changed expression in the rav2 knockout plants in the absence of HC-Pro as compared to wild type ( 15 of 433 were up-regulated; 98 of 433 were down-regulated ) . Together , these results suggest that HC-Pro causes major changes in global gene expression patterns , some of which are mediated by RAV2 . Interestingly , based on comparison of the set of genes with altered expression in rav2 mutant plants with the set altered by HC-Pro in a RAV2-dependent manner , it appears that HC-Pro changes the scope and spectrum of genes that are controlled by RAV2 . Gene Ontology ( GO ) term analysis was used to give a functional characterization of the tiling array results [52] . A key finding of this analysis was that multiple stress and defense responses were induced in HC-Pro expressing plants . The top four biological process categories that were over-represented among genes up-regulated in HC-Pro compared to wild type plants were: response to wounding ( 67 of 119 genes ) , response to jasmonic acid ( JA ) stimulus ( 48 of 119 genes ) , cold stress ( 49 of 197 genes ) and heat stress ( 33 of 109 genes ) ( Fig . 6C ) . Strikingly , genes in these same four categories were also over-represented among the genes that are up-regulated by HC-Pro in a RAV2-dependent manner ( Fig . 6D ) . Tables showing the specific genes that are up-regulated by HC-Pro in each of these GO categories , as well as the subsets that require RAV2 for HC-Pro up-regulation are in the Supplementary Tables ( Tables S12–15 ) . These results indicate that RAV2 plays a role in altered expression of stress and defense pathways in HC-Pro plants . Interestingly , FRY1 and CML38 , both of which have been implicated as suppressors of silencing [23] , [51] and are induced by HC-Pro in a RAV2-dependent manner ( Fig . 6B ) , have GO annotations of response to cold and wounding , respectively , suggesting a link between silencing and other stress and defense pathways .
It has been over a decade since the first plant viral suppressors of RNA silencing were reported [53] , [54] , [55] , providing an early clue that silencing serves as an anti-viral defense in plants and leading to the identification of many other such silencing suppressors [56] . However , the mechanisms by which these viral proteins manipulate silencing have remained largely elusive . Here we report the identification of a host protein , the transcription factor RAV2 , that is required for suppression of silencing mediated by two unrelated viral proteins , potyviral HC-Pro and carmoviral P38 . RAV2 is part of a gene family that comprises six members , two of which ( RAV1; At1g13260 and RAV2-like; At1g25560 ) are very closely related to RAV2 . Surprisingly , however , neither of these related genes is able to compensate for the loss of RAV2 with respect to suppression of silencing mediated by either HC-Pro or P38 . This result indicates that RAV2 provides a unique function in suppression of silencing . The identification of RAV2 as an important element in viral suppression of silencing provides a handle for identifying additional host partners and thereby unraveling the pathway of host involvement in that process . The discovery that plant viruses from many unrelated families encode suppressors of silencing has underscored the importance of silencing in antiviral defense . Similarly , we expect our finding that viral suppressors from two unrelated viruses have evolved independently to require RAV2 underscores the importance of host proteins in viral counter-defense . In addition , it suggests that RAV2 represents an effective and readily subverted control point – either for suppression of silencing in general or for a subset of suppressors with some mechanistic features in common . It will be interesting to see how general the requirement for RAV2 is in viral suppression of silencing . How could a transcription factor such as RAV2 be used to suppress silencing ? Two reports have identified RAV2 as a repressor of at least some target genes [57] , [58] . Therefore , it seemed reasonable to hypothesize that the role of RAV2 in HC-Pro suppression of silencing is to repress transcription of genes that encode components of the silencing machinery for the anti-viral branch of the silencing pathway . However , our global analysis of genome expression indicates that the expression of genes known to be required for RNA silencing is unchanged in HC-Pro transgenic plants as compared to wild type controls . Instead , our data shows that RAV2 is required for HC-Pro-mediated up-regulation of some stress and defense response genes . Earlier work showing that induction of both biotic and abiotic stresses interferes with RNA silencing induced by a viral amplicon in tobacco is consistent with a mechanism in which induction of other defense responses can divert the host from antiviral silencing [59] . The observation that RAV2 is induced by the ethylene defense pathway and is also required for viral suppression of silencing emphasizes the importance of crosstalk among defense pathways and supports the idea that RAV2 constitutes an important control point for the integration of defense responses during virus infection . One puzzle raised by the observation that HC-Pro , which is a cytoplasmic protein [60] , [61] , interacts with a host transcription factor is: How and where do the two proteins have the opportunity to meet ? Although HC-Pro has been shown to accumulate in nuclear inclusions in certain potyviral infections , it is thought that such inclusions represent storage of excess protein [61] . Thus , it seems more likely that HC-Pro and RAV2 interact in the cytoplasm . Sequestering transcription factors in the cytoplasm is a common mechanism used in eukaryotic organisms for controlling the activity of such proteins [62] , [63] . The interaction of HC-Pro with RAV2 in the cytoplasm could either reflect a direct involvement of RAV2 itself in suppression of silencing or interference by HC-Pro in the cellular control of RAV2 – either to block activation or promote inappropriate activation – thereby changing host gene expression in such a way that promotes suppression of silencing . Elucidating these issues , as well as examining whether P38 also physically interacts with RAV2 , is likely to be a fruitful area of research . Another particularly interesting aspect of our results is the differential requirement for RAV2 in suppression of different small RNA-mediated processes . Both HC-Pro and P38 suppress transitive silencing in the absence of RAV2; yet , both suppressors require RAV2 for suppression of target degradation via the activity of primary siRNAs . Furthermore , although HC-Pro requires RAV2 to block the activity of primary siRNAs , RAV2 is not required for HC-Pro-mediated defects in miRNA activity . Our present work does not distinguish whether these differential requirements for RAV2 indicate a fundamental difference in the mechanisms responsible for suppression of these processes or simply a difference in the cofactor requirements of a common mechanism . One current model for viral suppression of small RNA pathways posits a general mechanism in which small RNA duplexes are bound by the suppressor , thereby blocking the incorporation of one strand of the duplex into an active effector complex [17] , [64] . Our data showing a role for RAV2 in suppression of silencing does not directly support this proposed mechanism , but is also not inconsistent with it . Indeed , it has been shown that small RNA binding by HC-Pro in vitro is enhanced by unknown cellular factors [17] , [64] . Thus , RAV2 might be one such factor , acting either directly or indirectly to enhance small RNA binding . Expression of HC-Pro in transgenic plants causes a set of morphological anomalies that have been attributed to defects in the biogenesis and function of endogenous miRNAs [33] . However , there is emerging evidence that suggests that the phenotypic changes are largely independent of the miRNA pathway [14] , [15] , [20] , [65] . In support of this notion , the data we have presented here indicate that many of the HC-Pro-mediated morphological anomalies are RAV2-dependent whereas the defects in the miRNA pathway are RAV2-independent , arguing against a causative role for miRNAs in most HC-Pro-associated morphological anomalies Although the mechanism by which HC-Pro uses RAV2 to suppress silencing is not yet clear , the results of our tiling array analysis suggest two interesting , though speculative , possibilities . The first of these relates to the induction of AGO2 and a subset of other AGO genes in HC-Pro transgenic plants , an effect that is only partially dependent on RAV2 . The AGO genes that are up-regulated by HC-Pro are not required for post-transcriptional gene silencing ( PTGS ) . These results suggest that an alteration of the mix of AGO proteins in the cell might tip the balance away from PTGS towards other small RNA pathways that are not directly involved in anti-viral defense . The recent demonstration that changing the 5′ nucleotide of a miRNA so as to favor binding to AGO2 instead of AGO1 inactivates that miRNA [66] supports the idea that an overabundance of the wrong AGO proteins could contribute to suppression of silencing . The second interesting possibility suggested by our tiling data concerns the result that HC-Pro requires RAV2 to induce expression of FRY1 and CML38 , both of which have been implicated as endogenous suppressors of silencing and both of which are associated with stress or defense responses . Induction of endogenous suppressors of silencing may be more widespread than we know because most have probably not yet been identified [51] . It is tempting to speculate that the induction of stress and defense pathways by HC-Pro might have the counter-productive result - from the plant's perspective - of inducing a set of endogenous suppressors of antiviral silencing .
The tobacco 6b5 [32] and Arabidopsis TuMV HC-Pro [CT25 [14]] , TCV-P38 [39] , 306 and 6b4 [35] lines have been previously described . The Arabidopsis rav2/edf2 ( At1g68840 ) T-DNA insertion line ( SALK_070847 ) was used and did not express detectable levels of RAV2 mRNA as assayed by northern analysis . See Text S1 for the procedures used to generate the 35S:ntRAV tobacco transgenic line and genotyping of the SALK_070847 T-DNA insertion line . All Arabidopsis plants were of the Columbia ecotype . Histochemical staining for GUS activity was carried out as described [53] . The silencing of endogenous CH42 expression using the geminivirus CaLCV vector was performed exactly as described previously [14] . RNA isolation and RNA gel blot analysis of high and low molecular weight RNA were performed exactly as previously described [14] , [40] , [67] . Probes for detection of TuMV HC-Pro , TCV-P38 and 6b4 mRNAs , miRNA as well as those for primary and secondary siRNAs from the 6b4/306 transgene silencing system were previously described [7] . The RAV2 probe was generated using the primer set ( 5′ primer-TTGGAAAGTTCGGTCTGGTC and 3′ primer-TAATACGACTCACTATAGGGACCGCAAACATATCATCAACATCTC ) , which generate a 152 bp fragment from the 3′ end of the gene . The 3′ RAV2 primer contains T7 promoter sequences and a 4 nucleotide spacer at its 5′ end to facilitate synthesis of the probe using T7 polymerase . The production of the HC-Pro-GST fusion protein and 35S-methionine labeled ntRAV is described in Text S1 . To determine if HC-Pro-GST and ntRAV interact , approximately equimolar amounts of GST or HC-Pro-GST fusion protein were added to 20 µl of glutathione sepharose 4B beads ( GE Healthcare ) in GLB buffer ( 50 mM Tris-HCl , pH 8 . 0 , 150 mM NaCl , 1 mM EDTA , and 1 mM PMSF ) supplemented to contain 100 µg/ml BSA and 0 . 1% NP-40 ( Roche ) and shaken gently for 1 hour at 4°C . After rinsing with supplemented GLB , an equal amount of 35S-methionine labeled ntRAV was added to each sample , shaken gently at 4°C for 2 hours and rinsed again with supplemented GLB . Bound protein was eluted from the beads with Laemmli sample buffer , resolved by SDS-PAGE , and transferred to PVDF membrane . 35S-methionine labeled ntRAV was visualized by autoradiography . Protein was extracted from 0 . 5 g of Arabidopsis rosette leaf tissue by the following procedure . Tissue was frozen in liquid nitrogen , ground into powder with a mortar and pestle , homogenized in 4 ml of protein extraction buffer ( 40 mM Tris-Cl , pH 8 . 0 , 200 mM NaCl , 2 . 5 mM EDTA , 1% Triton X-100 , 0 . 1% NP-40 ) containing protease inhibitor cocktail ( Roche ) , and centrifuged ( 12 , 000 g at 4°C ) . The supernatant was incubated with 100 µl pre-washed anti-FLAG M2 agarose beads ( Sigma F2426 ) at 4°C for two hours . Agarose beads containing protein complexes were washed three times with extraction buffer , boiled in SDS sample buffer , resolved on a 10% SDS polyacrylamide gel , and subjected to western blotting . The presence of RAV2 protein was detected using a rabbit anti-RAV2 peptide antibody generated from the peptide GGKRSRDVDDMFALRC , and a rabbit anti-HC-Pro peptide antibody generated from the peptide KEFTKVVRDKLVGE was used to detect HC-Pro . Both RAV2 and HC-Pro peptide antibodies were produced by Sigma-Genosys . Total RNA was isolated as described above from the above ground portions of six week old plants that had not yet bolted . Generation of probes to poly-A RNA and hybridization to the tiling arrays were performed as described previously [46] , [47] . The data was analyzed using the program TileMap with a posterior probability of 0 . 8 [48] . The TileMap program identifies sequences that have significant changes in expression compared to controls , but does not provide fold-differences in expression levels . GO analysis was performed using ProfCom [68] . | RNA silencing is an important antiviral defense in plants , and many plant viruses encode proteins that block RNA silencing . However , the mechanism of action of the viral suppressors is complex , and little is known about the role of host plant proteins in the process . Here we report the first example of a host protein that plays a required role in viral suppression of silencing—a transcription factor called RAV2 that is required for suppression of silencing by two different and unrelated viral proteins . Analysis of plant gene expression patterns shows that RAV2 is required for induction of many genes involved in other stress and defense pathways , including genes implicated as plant suppressors of silencing . Overall , the results suggest that RAV2 is an important factor in viral suppression of silencing and that the role of RAV2 is to divert host defenses toward responses that interfere with antiviral silencing . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"biology/plant-biotic",
"interactions",
"virology/virulence",
"factors",
"and",
"mechanisms",
"molecular",
"biology/mrna",
"stability",
"plant",
"biology",
"virology",
"virology/mechanisms",
"of",
"resistance",
"and",
"susceptibility,",
"including",
"host",
"genetics... | 2010 | Two Plant Viral Suppressors of Silencing Require the Ethylene-Inducible Host Transcription Factor RAV2 to Block RNA Silencing |
A large proportion of age-related hearing loss is caused by loss or damage to outer hair cells in the organ of Corti . The organ of Corti is the mechanosensory transducing apparatus in the inner ear and is composed of inner hair cells , outer hair cells , and highly specialized supporting cells . The mechanisms that regulate differentiation of inner and outer hair cells are not known . Here we report that fibroblast growth factor 20 ( FGF20 ) is required for differentiation of cells in the lateral cochlear compartment ( outer hair and supporting cells ) within the organ of Corti during a specific developmental time . In the absence of FGF20 , mice are deaf and lateral compartment cells remain undifferentiated , postmitotic , and unresponsive to Notch-dependent lateral inhibition . These studies identify developmentally distinct medial ( inner hair and supporting cells ) and lateral compartments in the developing organ of Corti . The viability and hearing loss in Fgf20 knockout mice suggest that FGF20 may also be a deafness-associated gene in humans .
Congenital hearing loss is one of the most common hereditary diseases , affecting 2–3 infants per 1 , 000 live births [1] . Acquired age-related hearing loss affects one-third of people over the age of 65 [2] . A large proportion of age-related hearing loss is sensorineural and is caused by loss or damage to outer hair cells ( OHC ) in the organ of Corti ( OC ) [3] , [4] . The OC is the sensory transducing apparatus in the cochlea and is composed of one row of inner hair cells ( IHC ) and three rows of OHCs that are separated by two pillar cells ( PCs ) that form the tunnel of Corti . Each sensory hair cell is associated with an underlying supporting cell ( SC ) . Although there has been progress in understanding mechanisms of hair cell ( HC ) and SC differentiation [5] , [6] , the cellular signals that specify the distinct phenotypes of cochlear IHCs and OHCs are not known [7] . Fibroblast growth factor ( FGF ) signaling has essential functions at several stages of inner ear development . In the embryonic day 9–10 ( E9–E10 ) developing mouse , FGF3 , FGF8 , and FGF10 are essential for development of the otic vesicle [8] . These ligands signal through FGF receptor ( FGFR ) 2b in otic epithelium , and mice lacking Fgfr2b show impaired otic vesicle development [9] . At later stages of development , FGF signaling is required for morphogenesis of the organ of Corti . At E11 . 5 , Fgfr1 is expressed in the ventromedial wall of the otocyst , the region that will give rise to the cochlea [10] . At E15 , Fgfr1 expression is observed in the sensory epithelium of the developing cochlea [11] , [12] . Conditional disruption of Fgfr1 in sensory epithelial progenitor cells ( with Foxg1cre ) resulted in a severe reduction in HC number , possibly due to reduced proliferation of progenitor cells [10] . A similar phenotype was also observed in organ cultures treated with FGFR inhibitors [11] . Fgf20 is expressed in the presumptive epithelial domain of the developing cochlea at E13 . 5 and antibody inhibition of FGF20 in cochlear organ culture resulted in fewer SCs and HCs [11] . These studies suggest that FGF20 might be the ligand for FGFR1 during the early growth and differentiation stages of cochlear development . At later stages of organ of Corti development ( after E15 ) , inhibition of FGF signaling results in loss of PCs , suggesting an additional stage-specific role for FGF signaling [13] . Genetic and gene expression data suggest that this function is mediated by FGF8 signaling to FGFR3 . Fgfr3 is expressed after E15 . 5 in undifferentiated postmitotic progenitor cells that are thought to have the capacity to form OHCs , Deiters' cells ( DCs ) , PCs , and Hensen's cells ( HeCs ) [12]–[15] . Genetic disruption of Fgfr3 prevents the differentiation of PCs and the formation of the tunnel of Corti and results in deafness [13] , [15] , [16] . FGF8 is expressed in IHCs where it induces differentiation of PCs and formation of one row of OHCs through signaling to FGFR3 [10] , [17] , [18] . The mechanisms that regulate the formation of OHCs are particularly significant , given the importance of OHCs for hearing function and age-related hearing loss . Although mouse mutants lacking FGFR1 suggest a role for FGF signaling in OHC development [10] , the underlying mechanisms regulating OHC development are not known . Here we generated mice lacking FGF20 ( Fgf20βGal/βGal ) . We show that Fgf20βGal/βGal mice are viable , healthy , and congenitally deaf , specifically lack OHCs and outer supporting cells , and have patterning defects throughout most of the cochlear sensory epithelium . These studies show that the organ of Corti can be subdivided into developmentally distinct medial ( IHCs and inner SCs ) and lateral ( OHCs and outer SCs ) compartments that are under the control of distinct developmental programs . This model posits the existence of distinct progenitor cells that give rise to medial and lateral compartments of the OC .
To study the function of Fgf20 in vivo , we generated Fgf20 null mice in which exon 1 was replaced with a β-galactosidase gene ( Fgf20βGal ) ( Figure S1 ) . Homozygous Fgf20βGal/βGal mice were viable , fertile , and healthy . However , Fgf20βGal/βGal mice lacked auditory perception ( no ear twitch response to loud noise ) and had auditory brainstem response ( ABR ) thresholds greater than 40 db above controls in the 5–20 kHz range ( Figure 1A ) . Histological sections of the adult inner ear of Fgf20βGal/βGal mice showed normal gross morphology of the temporal bone and cochlea ( Figure S1E–G ) ; however , the OC showed significant dysmorphology , with variability in the degree of disorganization along the length of the cochlea ( Figure 1B ) . Some sections showed almost complete absence of sensory HCs and SCs , while other sections showed loss of OHCs and DCs . In contrast , wild type and heterozygous littermates showed normal cochlear organization , with one IHC , three OHCs , one inner and outer pillar cell ( IPC , OPC ) , and three DCs ( Figure 1B ) . To identify spatial and temporal patterns of Fgf20 expression in the developing inner ear , we stained whole mount preparations for β-galactosidase ( βGal ) activity . In Fgf20βGal/+ embryos , βGal was first detected in the anterio-ventral region of the otic vesicle at E10 . 5 , the region of the otic vesicle where sensory progenitor cells are located ( Figure 2A ) [19] . In histological sections of the otic vesicle , Fgf20-βGal was expressed within the domain of Sox2+ sensory progenitor cells at E11 . 5 ( Figure 2B ) . At E14 . 5 , the time of sensory cell specification , Fgf20-βGal was expressed in the Sox2+ , p27+ sensory domain ( Figure 2C , D ) , in an apical to basal graded expression pattern , similar to previously reported expression patterns for Fgf20 [11] . At postnatal day 0 ( P0 ) , a time when almost all sensory cells have completed differentiation , Fgf20-βGal was expressed throughout all inner ear sensory epithelia ( Figure 2E and Figure S2 ) . In the cochlea , Fgf20-βGal expression was restricted to SCs and was expressed in a graded medial to lateral pattern , with highest levels in the inner phalangeal cells ( IPhC ) and lower levels in PCs ( Figure 2F and Figure S2A ) . Fgf20-βGal was also expressed in the vestibular sensory organs of the inner ear , including the maculae of the utricle and saccule and cristae of the semicircular canals ( Figure S2B–F ) . Fgf20βGal/βGal mice did not show any vestibular dysfunction ( unpublished data ) . To further analyze the hair cell phenotype in the cochlea , we isolated whole cochleae at P0 and stained with phalloidin , as well as with antibodies against Myo6 and Calretinin ( Figure 3 and Figure S3 ) [20] , [21] . Wild type cochleae showed three rows of OHCs and one row of IHCs throughout the OC ( Figure 3A ) . However , in the cochlea of Fgf20βGal/βGal newborn pups , the proximal base region contained only two rows of OHCs and one row of IHCs . In the middle and apical regions , patches of HCs were observed ( Figure 3B ) . Such patches typically contained three rows of OHCs and two rows of IHCs , and there were no HCs in the regions between the patches . Finally , no distinct phalloidin or Myo6 positive HCs were present in the most apical region . Because HC differentiation progresses from the base to the apex of the cochlea , we sought to determine whether differentiation of HCs in the distal apex region of Fgf20βGal/βGal cochlea was delayed or whether HCs were lost . At P7 , expression of HC markers in the distal apex of Fgf20βGal/βGal and Fgf20βGal/+ cochlea were comparable ( Figure 3 and Figure S3 ) , indicating that HCs were not lost but rather delayed in differentiation . Consistent with delayed HC differentiation , at E16 . 5 , HCs were undifferentiated in the middle region of the cochlea of Fgf20βGal/βGal compared to Fgf20βGal/+ cochlea ( Figure S3A , B ) . To identify whether SCs were properly formed , we stained P0 cochlea with antibodies against Prox1 and p75 . At P0 , Prox1 is expressed at high levels in DCs and PCs , while p75 labels PCs and HeCs [22] , [23] . In Fgf20βGal/+ cochleae , there were two rows of PCs , three rows of DCs , and one row of HeCs ( Figure S3G ) . Immunolabeling of Fgf20βGal/βGal cochleae showed that all the SC types existed , although with dysmorphology . Similar to HC patterns , two rows of DCs and two rows of PCs were formed at the base region of Fgf20βGal/βGal cochlea ( Figure S3H ) . In the middle region , where HCs were clustered in patches , SCs were formed in accordance with the hair cell pattern ( three rows of DCs and two rows of PCs ) . No SCs were observed in the space between the sensory patches . Interestingly , within a patch , OHCs were surrounded by PCs and HeCs , as indicated by continuous p75 staining ( Figure S3J ) . Unlike HCs , apical SCs were differentiated , as indicated by Prox1 staining ( Figure S3G , H ) , suggesting that apical SCs may develop normally in the absence of FGF20 . Because of the delay in HC differentiation and patch formation , the total numbers of cochlear HCs were quantified at P4 , a time when both Fgf20βGal/+ and Fgf20βGal/βGal cochlea appeared fully differentiated ( Figure S3K ) . Surprisingly , although there were two rows of IHCs in patches and no IHCs in the region between patches , the total number of IHCs was the same in Fgf20βGal/+ and Fgf20βGal/βGal cochlea ( 680±18 , n = 4 in Fgf20βGal/+ and 685±58 , n = 3 in Fgf20βGal/βGal , p = 0 . 4 ) . However , the number of OHCs was decreased by 70% in Fgf20βGal/βGal compared to Fgf20βGal/+ cochlea ( 2035±42 in Fgf20βGal/+ and 601±60 in Fgf20βGal/βGal , p<0 . 001 ) ( Figure 3D ) . In addition , cochlear length was decreased by 10% in Fgf20βGal/βGal compared to Fgf20βGal/+ mice ( Figure 3E ) . We also counted the number of SCs ( DCs , OPCs , and IPCs ) normalized to 100 µm intervals . Similar to the large decrease in the number of OHCs , the number of DCs+OPCs was decreased by 52% in Fgf20βGal/βGal compared to Fgf20βGal/+ cochlea ( 55±3 , n = 6 in Fgf20βGal/+ and 26±4 , n = 8 in Fgf20βGal/βGal per 100 µm , p<0 . 001 ) ( Figure 3F ) . In contrast , the number of IPCs was decreased by only 15% in Fgf20βGal/βGal compared to Fgf20βGal/+ cochlea ( 19±1 in Fgf20βGal/+ and 16±1 in Fgf20βGal/βGal per 100 µm , p<0 . 01 ) ( Figure 3F ) . Next , we compared the ratio of different cell types in Fgf20βGal/βGal and Fgf20βGal/+ cochlea . In Fgf20βGal/+ cochlea , the ratio of OHC/IHC was 3 . 4±0 . 3 . However , in Fgf20βGal/βGal cochlea this ratio was decreased ( by 62% ) to 1 . 3±0 . 6 ( p<0 . 001 ) . Additionally , the ratio of DC+OPC/IPC in Fgf20βGal/+ cochlea was 2 . 9±0 . 1 , but was decreased ( by 45% ) to 1 . 6±0 . 2 in Fgf20βGal/βGal cochlea ( p<0 . 001 ) . Interestingly , within the lateral and medial compartments , the ratio of DC+OPC/OHC ( 1 . 4±0 . 0 in Fgf20βGal/+ and 1 . 5±0 . 1 in Fgf20βGal/βGal , p<0 . 01 ) and IPC/IHC ( 1 . 6±0 . 1 in Fgf20βGal/+ and 1 . 2±0 . 1 in Fgf20βGal/βGal , p<0 . 001 ) was slightly increased ( by 7% ) or decreased ( by 25% ) , respectively , in Fgf20βGal/βGal compared to Fgf20βGal/+ cochlea . These ratios indicate that absence of FGF20 primarily affects lateral compartment cells ( i . e . , OHCs and DCs ) . Next we asked whether loss of the lateral compartment was due to loss of sensory domain progenitor cells . To do this , we labeled E13 . 5 or E14 . 5 cochlea for Sox2 , a marker for sensory progenitors , or Jag1 , a marker for Kölliker's organ [24] , [25] . The expression pattern of Sox2 and Jag1 was comparable in Fgf20βGal/βGal and Fgf20βGal/+ cochlea at E13 . 5 and E14 . 5 ( Figure S4A–F ) . Additionally , cell proliferation was comparable in Fgf20βGal/βGal and Fgf20βGal/+ E13 . 5 cochlea ( Figure S4G , H ) , indicating that the sensory domain had formed normally . Next , we hypothesized that FGF20 may play a role in lateral compartment differentiation . To test this , we isolated E13 . 5 or E14 . 5 cochlea and treated explants with 1 µM FGF9 ( which shows similar biochemical activity in vitro compared to FGF20 ) [26] beginning at E13 . 5 , E14 . 5 , E15 . 5 , and E16 . 5 . Control cultures were maintained in parallel , but did not receive FGF9 . Cochlear explants were cultured in these media for 5 d and then stained with Myo7a antibodies ( to identify HCs ) [20] , [27] . Untreated Fgf20βGal/+ explants showed normal patterning , with one row of IHCs and 3–4 rows of OHCs ( Figure 4A , E and Figure S5A , E ) . Also , untreated Fgf20βGal/βGal explants showed the expected patterning defects ( patches of HCs and SCs towards the apical cochlea ) and loss of HCs ( Figure 4C , G and Figure S5C , G ) . Notably , however , treatment of Fgf20βGal/βGal explants with FGF9 at E13 . 5 and E14 . 5 resulted in rescue of the cochlear phenotype , with the cochlea showing a normal and contiguous pattern of sensory cells and increased numbers of OHCs ( 162±53 , n = 2 without FGF9 and 486±122 , n = 3 with FGF9 treatment at E13 . 5 , p<0 . 05 , and 376±96 , n = 3 without FGF9 and 725±100 , n = 3 with FGF9 treatment at E14 . 5 , p<0 . 01 ) compared to untreated explants ( Figure 4C , D , I and Figure S5C , D , I ) . Finally , treatment of Fgf20βGal/βGal explants with FGF9 at E15 . 5 or E16 . 5 did not affect the number of HCs ( 168±0 , n = 2 without FGF9 and 176±90 , n = 3 with FGF9 treatment at E15 . 5 , p = 0 . 9 , and 238±40 , n = 4 without FGF9 and 279±47 , n = 4 with FGF9 treatment at E16 . 5 , p = 0 . 3 ) ( Figure 4G , H , I and Figure S5G , H , I ) , indicating that normal differentiation of lateral compartment cells requires active FGF signaling prior to E14 . 5 . Patterning and numbers of SCs were also rescued by FGF9 treatment ( 640±49 , n = 3 without FGF9 and 1001±39 , n = 4 with FGF9 treatment at E14 . 5 , p<0 . 001 ) ( Figure 4J–N ) . BrdU labeling of these cultures indicated that treatment of Fgf20βGal/βGal cochlea with FGF9 did not induce renewed proliferation within the sensory epithelia ( Figure S5J and K ) at this stage , which indicated that FGF9 treatment functioned to induce lateral compartment cell differentiation into HCs and SCs . Also , treatment of Fgf20βGal/+ explants with FGF9 did not change the morphology or number of HCs or SCs , indicating that FGF signaling is not sufficient to induce ectopic HC or SC formation ( Figure 4B , F , I , K and Figure S5B , F , I ) . Similar experiments were repeated with recombinant FGF20 protein with qualitatively similar results , although FGF20 was less active than FGF9 in this assay ( unpublished data ) . To determine whether the missing outer compartment cells were lost or were still present in an undifferentiated state , we stained whole cochleae of P0 pups with E-Cadherin antibodies , which marks lateral compartment cells at late gestational and postnatal stages of development [6] . In Fgf20βGal/+ cochlea , E-Cadherin labeled all lateral compartment cells including OHCs , DCs , and HeCs ( Figure 5A , upper ) . Interestingly , in Fgf20βGal/βGal cochlea , E-Cadherin was highly expressed in the region between the sensory patches where there were no HCs or SCs ( Figure 5A , lower ) , identifying these as potential lateral compartment cells . We also labeled specimens for Sox2 and with phalloidin . At P0 , Sox2 labels all supporting cells , but at earlier stages of sensory domain formation ( E14 . 5 ) , Sox2 labels undifferentiated sensory cells [24] . We observed normal patterns of Sox2 expression in Fgf20βGal/+ cochlea . However , in Fgf20βGal/βGal cochlea , Sox2 was expressed both in supporting cells and in the regions between the sensory patches ( Figure 5B arrows ) . We also examined the expression of p27 , a marker of SCs and undifferentiated sensory progenitors [28] . The expression pattern of p27 was similar to that of Sox2 , with high expression in cells in the region between the sensory patches ( Figure 5C ) . Although the identity of the cells within these gaps in the sensory epithelium is not known , these expression studies suggest that these cells may be an arrested progenitor-like cell or a differentiated non-sensory cell . To determine whether the lineage precursors of these cells could be rescued , E14 . 5 Fgf20βGal/βGal cochlea explants were treated with or without FGF9 and co-stained for Sox2 and Prox1 expression after 5 d in culture . In explants not exposed to FGF9 , the region between the patches was Sox2+; Prox1− ( Figure 5D , left ) . However , following exposure to FGF9 , these cells became Sox2+; Prox1+ ( Figure 5D , right ) , indicating that the lineage precursors of these cells are undifferentiated sensory cells and that exposure to FGF9 induced their differentiation into lateral HCs and SCs , such as OHCs and DCs . This finding indicates that FGF signaling is required to induce differentiation of cells in the lateral cochlear compartment . Next , we asked whether FGF20 functions to induce differentiation of a specific cell phenotype in the lateral compartment versus functioning as a gate to permit lateral compartment differentiation . To answer this question , we treated E14 . 5 cochlea explants with DAPT , a γ-secretase inhibitor , which inhibits the Notch signaling pathway [29] . At this stage of development , the Notch pathway functions to prevent SC differentiation into HCs [30] . In Fgf20βGal/+ explants treated with DAPT , the domain of IHCs and OHCs expanded at the expense of SCs , compared to untreated explants ( Figure 6A and B ) . In Fgf20βGal/βGal explants , treatment with DAPT also expanded the IHC domain , similar to heterozygous explants treated with DAPT . However , the domain of OHCs in DAPT-treated Fgf20βGal/βGal explants was still smaller than the OHC domain of DAPT-treated Fgf20βGal/+ explants and also contained patches of undifferentiated sensory progenitor cells ( Figure 6C and D ) . This finding indicates that DAPT treatment did not induce differentiation of otherwise FGF20-dependent precursor cells . Also , we observed a dramatic reduction of SCs following DAPT treatment of either genotype , indicating that all SCs were converted into HCs ( Figure 6F and H ) . Together , these data support a model in which FGF20 functions as a permissive factor that is required to initiate lateral compartment differentiation before E14 . 5 ( Figure 7 ) . Without FGF20 , lateral sensory cells remained in an undifferentiated state .
To identify the developmental time when FGF20 functions to regulate lateral compartment differentiation , rescue experiments were performed in which Fgf20βGal/βGal cochlea were placed in culture prior to differentiation ( E13 . 5 ) and then FGF9 was added to the culture at different time points . These experiments showed that the lateral compartment differentiation defects in Fgf20βGal/βGalmice could only be rescued if FGF9 was added at or before E14 . 5 . However , treatment with FGF9 at or after E15 . 5 failed to rescue the phenotype of Fgf20βGal/βGal mice . This is interesting because E14 . 5–15 . 5 corresponds to the time when sensory cell specification is completed and HC and SC differentiation begins [5] , [32] . The changes in the cochlear epithelium that renders it non-responsive to FGF signaling after E14 . 5 are not known . Possibilities include loss of FGFR1 expression , uncoupling of FGFR1 to cellular signal transduction pathways , or loss of cofactors required for ligand activation of FGFR1 . In lung development , a feed forward signaling loop couples FGF9 with Wnt/β-catenin signaling and maintenance of FGFR expression . Loss of Fgf9 resulted in loss of Fgfr1 and Fgfr2 expression and subsequent loss of responsiveness of explanted lung to exogenous FGF9 [33] . If a similar feed forward loop functions in the inner ear prosensory epithelium , loss of FGFR expression in Fgf20βGal/βGal mice could explain the loss of responsiveness to exogenous FGF after E14 . 5 . However , in the inner ear , FGF20 continues to be expressed in IPhCs and at low levels in PCs until early postnatal ages ( Figure 2 ) . This suggests that FGF20 signaling may have additional roles in cochlear development . At P0 , βGal staining indicates that Fgf20 is expressed at highest levels in the cochlear apex ( Figure 2 ) . Since differentiation of the apical cochlea is delayed in Fgf20βGal/βGal mice ( Figure 3 ) , FGF20 may function at later stages of development to promote sensory cell maturation . Because damage or loss of OHCs is thought to be a major cause of sensorineural hearing loss , efforts to restore hearing in some patients with sensorineural hearing loss will require regeneration of OHCs . Understanding the changes that occur in sensory progenitor cells between E14 . 5 and E15 . 5 is important because they may provide clues about pathways required for reactivation of OHC progenitors in the adult or protecting OHCs from ototoxic damage . Although FGF20 signaling alone may not be sufficient to induce regeneration of OHCs , it may be required in combination with other signaling molecules . For example , in lung development , responsiveness of lung tissue lacking FGF9 can be restored by simultaneously treating Fgf9−/− explants with activators of Wnt/β-catenin signaling and with FGF9 [34] . The Fgf9 subfamily includes Fgf16 and Fgf20 [35] . Consistent with the conserved sequences within this subfamily , the biochemical activities of FGF20 are similar to that of FGF9 and FGF16 [26] . In vitro , FGF20 binds and activates c splice variants of FGFR1 , FGFR2 , and FGFR3 , which are generally expressed in mesenchymal cells , and b splice variants of FGFR3 , which are expressed in epithelial cells [26] . However , the phenotype of Fgf20βGal/βGal mice is most similar to that of Fgfr1 conditional deletion mutants , in which epithelial Fgfr1 was inactivated in the developing inner ear sensory epithelium with Foxg1cre [10] . The phenotypic similarities strongly suggest that FGFR1 is the physiological receptor for FGF20 . Because , in vitro , FGF20 activates FGFR1c to a much greater extent than FGFR1b [26] , the FGFR1c variant may be expressed in the developing cochlear sensory epithelium . Alternatively , unique cofactors within the cochlear sensory epithelium may allow FGF20 to activate FGFR1c . The sensory epithelium of the mammalian cochlea cannot regenerate following ototoxic or noise damage; however , the avian and amphibian inner ear responds to ototoxic or noise-induced injury with a robust regenerative response that results in complete functional recovery [36] . The underlying mechanisms accounting for this difference in regenerative capacity are not understood . However , in principal , therapeutic reactivation of appropriate signaling pathways in the mammalian inner ear should be able to recapitulate the avian response , resulting in both functional repair and prevention of further pathology . Our observation that FGF20 functions as a permissive factor for lateral compartment differentiation suggests that FGF signaling may be a necessary factor for promoting inner ear regeneration . Additionally , zebrafish lacking FGF20 are viable and healthy , but have defects in their ability to regenerate damaged fins [37] . These observations suggest that FGF signaling , and specifically FGF20 or related FGFs , may be important factors for regeneration of a variety of tissues , including the inner ear . Inducible genetic systems in the mouse and the identification of signaling pathways that interact with FGF20 will be required to test the protective or regenerative potential of FGF20 in noise or ototoxic damaged mammalian inner ear .
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 Washington University Division of Comparative Medicine Animal Studies Committee ( Protocol Number 20100223 ) . All efforts were made to minimize animal suffering . The Fgf20 targeting construct was made using recombineering methods as previously reported [38] . Briefly , exon1 of Fgf20 was replaced with a βGal–LoxP-neomycin-LoxP cassette to generate Fgf20βGal ( neo ) /+ mice . The neomycin gene was eliminated by mating with β-actincre mice to generate Fgf20βGal/+ mice . Fgf20βGal/+ males and females were crossed to generate Fgf20βGal/βGal mice . Genotyping was performed using PCR1: CTGCATTC GCCTCGCCACCCTTGCTACACT; PCR2: GGATCTGCAGGTGGAAGCCGGTGCGGCAGT; PCR3: GGCCTTCCTGTAGCCAGCTTTCATCAACAT primers , which amplify wild type ( 335 bp ) and mutant ( 498 bp ) PCR fragments as indicated in Figure S1A . Mice were maintained on a 129X1/SvJ;C57B6/J mixed background . Fgf20βGal/+ and Fgf20βGal/βGal mice were viable and fertile . Mice were anesthetized by i . p . administration of ketamine ( 80 mg/kg ) and xylazine ( 15 mg/kg ) , and maintained at 37°C throughout the testing . ABR testing was carried out in a single walled sound-attenuating room . Testing was similar to what has been previously described [39] . Briefly , stimulus presentation and data acquisition were performed with TDT System 3 equipment using SigGen and BioSig software ( Tucker Davis Technologies ) . An ES-1 electrostatic speaker was placed 7 cm from the animal's right ear . Toneburst stimuli ( 5 , 10 , 20 , 28 , and 40 kHz ) were 5 ms in length with a 1-ms rise/fall time . Stimuli were presented at decreasing intensities in 5 dB steps until Wave I was no longer observed . Auditory profiles were recorded using platinum subdermal needle electrodes ( Grass Technologies ) placed with the recording electrode behind the right pinna , the reference electrode at the vertex , and ground electrode in the skin of the back . Responses were amplified and filtered ( X 100 , 000 and low filter: 100 Hz , high filter: 3 , 000 Hz ) using a Grass P 55 preamp . Tonebursts at each frequency and intensity were presented 1 , 000 times . Stimulus levels were calibrated using SigCal ( Tucker Davis Technologies ) program with an ACO Pacific ¼ inch microphone placed where the mouse's ear would be located . For detection of Fgf20 mRNA , E14 . 5 embryos were dissected and inner ear tissue was isolated . RNA was extracted with the RNeasy kit ( Qiagen ) . cDNA synthesis used the SuperScript III First-Strand Synthesis System for reverse transcription–polymerase chain reaction ( Invitrogen ) , following the manufacturer's protocol . Mouse Fgf20 mRNA levels were quantified using a TaqMan gene expression assay ( ABI Mm00748347_m1 ) . TaqMan assays were run in an ABI7500 fast real-time PCR machine . Cochleae were dissected from P0 pups in PBS and fixed overnight in Mirsky's Fixative ( National Diagnostics ) . For whole mount staining , samples were washed three times in PBT ( PBS , 0 . 1% Tween-20 ) and incubated in βGal staining solution ( 2 mM MgCl2 , 35 mM potassium ferrocyanide , 35 mM potassium ferricyanide , 1 mg/mg X-Gal in PBT ) at 37°C until color reaction was apparent . Samples were washed in PBS fixed in 10% formalin and imaged under a dissecting microscope . For staining histological sections , samples were cryosectioned , washed with PBS , and incubated in βGal staining solution . Samples were washed in PBS , embedded , and photographed . For adult histology , mice were sacrificed with an overdose of pentobarbital ( 200 mg/kg ) . Temporal bones were dissected free from the skull and broken in half to expose the cochlea , which was perfused via the round window with a solution containing 4% paraformaldehyde and 0 . 1% glutaraldehyde . Cochleae were further fixed in this solution overnight and then rinsed free of aldehyde with several changes of PBS . They were then post-fixed in 1% osmium tetroxide ( 30 min ) and rinsed and dehydrated through a series of acetones . Tissues were then infiltrated and embedded in an epon-araldite mixture and polymerized overnight at 60°C . Ten adjacent , 4 µm thick sections were saved from the midmodiolar plane from each cochlea , counterstained with toluidine blue , and coverslipped . For frozen sections , embryos were fixed with 4% paraformaldehyde overnight and washed with PBS . Samples were soaked in 30% sucrose and embedded in OCT compound ( Tissue-Tek ) . Samples were sectioned ( 12 µm ) and stored at −80°C for immunohistochemistry . For total hair cell counting , P4 cochleae stained with phallodin were used because both Fgf20βGal/+ and Fgf20βGal/βGal cochleae were completely differentiated at this stage . For supporting cell counting , P0 cochleae stained with either phallodin or Prox1 were used . Because of the incomplete staining pattern of Prox1 from the base to the apex , we could not count all of the supporting cells . Instead , we counted more than 300 µm regions of the base , middle , and apex of the cochlea and normalized counts to 100 µm . Inner and outer hair cells were identified by location and morphology of phalloidin staining . Inner pillar cells were distinguished by location and morphology among Prox1+ cells . Deiters' cells and outer pillar cells were counted by exclusion of inner pillar cells from Prox1+ cells . Cell counting was performed using Image J software . Embryonic mouse cochlear cultures were established as described previously [40] with minor modifications . In brief , cochleae from Fgf20βGal/+ and Fgf20βGal/βGal embryos at various ages ( E13 . 5–E16 . 5 ) were dissected , to expose the sensory epithelium , in ice-cold M199 Hanks solution , transferred to a Ma-Tek dish ( Ma-Tek Corporation ) , coated with Matrigel ( BD Biosciences ) , and maintained at 37°C in vitro in experimental ( FGF9 or DAPT ) or control culture media for 3–6 d . Recombinant FGF9 and FGF20 protein was obtained from PeproTech Inc . DAPT was obtained from Sigma . To activate FGF signaling , FGF9 culture media ( 1 µg/ml FGF9+1 mM Heparin in MEM+10% FBS ) was added to explant cultures at E13 . 5 , E14 . 5 , E15 . 5 , or E16 . 5 for 6 , 5 , 4 , or 3 d , respectively ( until age of E19 . 5 ) . Control culture media contained ( 1 µg/ml heparin in MEM+10% FBS ) . To inhibit Notch signaling , DAPT ( N-[ ( 3 , 5-Difluorophenyl ) acetyl]-L-alanyl-2-phenyl]glycine-1 , 1-dimethylethyl ester ) culture media was added to explant cultures at E14 . 5 for 5 d ( E19 . 5 ) . DAPT media: 10 µM DAPT ( reconstituted in DMSO ) in MEM+10% FBS . Control culture media contained DMSO in MEM+10% FBS . Cochleae were treated in pairs ( i . e . , cochlea from left ear received experimental media while the cochlea from the right ear of the same embryo received control media ) . The appropriate culture media was replaced every 24 h for all explants . Following incubation , explants were fixed in 4% paraformaldehyde for 30 min and analyzed by immunohistochemistry . Immunohistochemistry was described previously [41] . Briefly , for whole mount immunofluorescence , cochleae were isolated and fixed in 4% PFA overnight at 4°C . Samples were washed with PBS and blocked with PBS containing 0 . 1% triton X-100 and 0 . 5% goat serum . Primary antibody was incubated overnight at 4°C . Samples were washed with PBS and incubated with a secondary antibody for 2 h at room temperature . Samples were washed , placed on a glass microscope slide , coverslipped , and photographed using a Zeiss LSM 700 confocal microscope . For section immunofluorescence , frozen sections ( 12 µm ) were washed with PBS and blocked with 0 . 1% triton X-100 and 0 . 5% donkey serum . Sections were incubated with primary antibodies in a humidified chamber overnight at 4°C . Sections were then washed and incubated with secondary antibody for 1 h at room temperature . Samples were washed , coverslipped with Vectashield Mounting Media ( Vector lab ) , and photographed using a Zeiss LSM 700 confocal microscope . Primary antibodies used: Phallodin ( R&D Systems , 1∶40 ) , Myo6 ( Proteus Biosciences , 1∶500 ) , Myo7a ( Proteus Biosciences , 1∶500 ) , Calretinin ( Millipore , 1∶500 ) , Prox1 ( Covance , 1∶500 ) , p27 ( Neomarkers , 1∶500 ) , p75 ( Chemicon , 1∶500 ) , β-Galactosidase ( Abcam , 1∶500 ) , Sox2 ( Millipore , 1∶500 , Santa Cruz 1∶200 ) , BrdU ( BD Biosciences , 1∶500 ) , E-cadherin ( Invitrogen , 1∶500 ) , and Jag1 ( Santa Cruz 1∶200 ) . Number of samples is indicated for each experiment . All data are presented as mean ± standard deviation ( sd ) . The p value for difference between samples was calculated using a two-tailed Student's t test . p<0 . 05 was considered as significant . | A large proportion of age-related hearing loss is caused by loss or damage to outer hair cells in the organ of Corti . The organ of Corti is a highly specialized structure in the inner ear that is composed of inner hair cells , outer hair cells , and associated supporting cells . Although we understand some of the mechanisms that regulate hair cell versus supporting cell differentiation , the mechanisms that regulate differentiation of inner versus outer hair cells are not known . One potential candidate is fibroblast growth factor ( FGF ) signaling , which is known to regulate the morphogenesis of many sensory organs , including the organ of Corti . In this study , we find that FGF20 signaling is required at a specific time during development to initiate differentiation of cells in the mouse lateral cochlear compartment ( which contains outer hair cells and supporting cells , but not inner hair cells ) . In the absence of FGF20 , mice are deaf , and lateral compartment cells remain undifferentiated and unresponsive to mechanisms that regulate the final stages of differentiation . These findings are significant given the importance of outer hair cells during age-related hearing loss . Our studies also suggest that genetic mutations in FGF20 may result in deafness in humans and that FGF20 may be an important factor for the repair or regeneration of sensory cells in the inner ear . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"molecular",
"neuroscience",
"neurogenesis",
"animal",
"genetics",
"genetic",
"mutation",
"anatomy",
"and",
"physiology",
"neuroscience",
"cell",
"differentiation",
"dna",
"transcription",
"animal",
"models",
"developmental",
"biology",
"model",
"organisms",
"... | 2012 | Differentiation of the Lateral Compartment of the Cochlea Requires a Temporally Restricted FGF20 Signal |
Drug combinations for the treatment of leishmaniasis represent a promising and challenging chemotherapeutic strategy that has recently been implemented in different endemic areas . However , the vast majority of studies undertaken to date have ignored the potential risk that Leishmania parasites could develop resistance to the different drugs used in such combinations . As a result , this study was designed to elucidate the ability of Leishmania donovani to develop experimental resistance to anti-leishmanial drug combinations . The induction of resistance to amphotericin B/miltefosine , amphotericin B/paromomycin , amphotericin B/SbIII , miltefosine/paromomycin , and SbIII/paromomycin was determined using a step-wise adaptation process to increasing drug concentrations . Intracellular amastigotes resistant to these drug combinations were obtained from resistant L . donovani promastigote forms , and the thiol and ATP levels and the mitochondrial membrane potential of the resistant lines were analysed . Resistance to drug combinations was obtained after 10 weeks and remained in the intracellular amastigotes . Additionally , this resistance proved to be unstable . More importantly , we observed that promastigotes/amastigotes resistant to one drug combination showed a marked cross-resistant profile to other anti-leishmanial drugs . Additionally , the thiol levels increased in resistant lines that remained protected against the drug-induced loss of ATP and mitochondrial membrane potential . We have therefore demonstrated that different resistance patterns can be obtained in L . donovani depending upon the drug combinations used . Resistance to the combinations miltefosine/paromomycin and SbIII/paromomycin is easily obtained experimentally . These results have been validated in intracellular amastigotes , and have important relevance for ensuring the long-term efficacy of drug combinations .
The use of drug combinations , either in co-formulations or co-administrations , is an established approach for the treatment of several infectious diseases including malaria and tuberculosis [1] . This approach has also recently become a priority for other tropical parasitic diseases , such as visceral leishmaniasis [2]–[6] . Leishmaniasis , a neglected tropical parasitic disease that is prevalent in 98 countries spread across three continents , is caused by protozoan parasites belonging to the genus Leishmania [7] . The estimated incidence of leishmaniasis is 0 . 2–0 . 4 million cases of the visceral form ( VL ) and 0 . 7–1 . 2 million cases of the cutaneous form ( CL ) [7] . Although chemotherapy is the only current treatment option for leishmaniasis , its efficacy is increasingly limited by growing resistance to first-line drugs , especially antimonials , the frequent side-effects associated with their use , and the high cost of treatment [7] , [8] . The recommended first-line therapies for VL include: i ) pentavalent antimonials ( meglumine antimoniate and sodium stibogluconate ) , except in some regions in the Indian subcontinent where there are significant areas of drug resistance [9]; ii ) the polyene antibiotic amphotericin B ( AmB ) ; iii ) the liposomal formulation AmBisome; iv ) the aminoglycoside paromomycin ( PMM ) ; and v ) the oral drug miltefosine ( MLF ) . Recently , the WHO [7] , [10] , recommended to use either a single dose of AmBisome or combinations of anti-leishmanial drugs in order to reduce the duration and toxicity of treatment , prolong the therapeutic life span of existing drugs and delay the emergence of resistance . Although recent clinical trials have highlighted the efficacy and safety of anti-leishmanial drug combinations [4] , [5] , [10]–[12] , additional clinical studies are needed to investigate various other factors , such as the identification of an effective , well-tolerated and short treatment regimen , logistical aspects , and the potential risk of developing resistance considering that compliance in field conditions can be low [13] . Herein we describe the selection and characterization of experimental resistance to drug combinations in Leishmania parasites . Our findings clearly demonstrate the acquisition of resistance to different drug combinations in Leishmania donovani promastigotes using a step-wise adaptation process to increasing drug concentrations . Similarly , and perhaps importantly , we have obtained intracellular L . donovani amastigotes that are resistant to different drug combinations from promastigote forms resistant to these same combinations . These results indicate different patterns of resistance depending on the drug combinations used , with the combination MLF/PMM selecting resistant L . donovani more rapidly than the combination AmB/PMM . Significantly , we have also observed that promastigotes/amastigotes resistant to one drug combination show a marked cross-resistance profile to other anti-leishmanial drugs , a finding that could be of major clinical relevance . Additionally , our results indicate that the resistant lines remain protected against the drug-induced loss of ATP and mitochondrial membrane potential .
Trivalent antimony ( SbIII ) , paromomycin ( PMM ) , amphotericin B ( AmB ) , paraformaldehyde , MTT [3- ( 4 , 5-dimethylthiazol-2-yl ) -2 , 5-diphenyltetrazolium bromide] , Rhodamine 123 ( Rh123 ) , buthionine sulfoximine ( BSO ) , FCCP ( carbonyl cyanide 4-trifluoromethoxyphenylhydrazone ) , and Triton X-100 were obtained from Sigma-Aldrich ( St . Louis , MO ) . Miltefosine ( MLF ) was purchased from Zentaris GmbH ( Frankfurt am Main , Germany ) , and CellTiter-Glo , CellTracker , and 4′ , 6-diamidino-2-phenylindole dihydrochloride ( DAPI ) were purchased from Invitrogen . L-glutamine and penicillin/streptomycin were obtained from Gibco . All chemicals were of the highest quality available . The L . donovani promastigotes ( MHOM/ET/67/HU3 ) and derivative lines used in this study were grown at 28°C in RPMI 1640-modified medium ( Invitrogen ) supplemented with 20% or 10% heat-inactivated fetal bovine serum ( HIFBS , Invitrogen ) . For thiol assays , they were grown in M-199 medium ( Gibco ) supplemented with 10% HIFBS . The resistant lines were obtained following a previously described step-wise adaptation process [14] , [15] . This process started with drug pressure in the wild-type ( WT ) L . donovani line at a concentration below the drug EC50 ( the concentration of the drug required to inhibit parasite growth by 50% ) , gradually increasing the drug pressure over 10 weeks . After this period , the resistant lines were maintained for eight further weeks at the final drug concentration . The drug combination resistant lines generated , based on WHO recommendations [7] , were AmB+MLF ( AM ) , AmB+PMM ( AP ) , AmB+SbIII ( the antimonial active form; AS ) , MLF+PMM ( MP ) and SbIII+PMM ( SP ) . Singly resistant lines named A , M , P , and S were obtained in a similar manner . All resistant lines were maintained in the continuous presence of drugs . Resistance stability was checked at one and four months after removal from drug pressure . The EC50 , resistance index ( EC50 ratio for resistant and WT parasites ) , and cross-resistance profile were determined for each line using an MTT colorimetric assay after incubation for 72 h at 28°C in the presence of increasing concentrations of the drug , as described previously [16] . Six-week-old male BALB/c mice were purchased from Charles River Breeding Laboratories and maintained in the Animal Facility Service of our Institute under pathogen-free conditions . They were fed a regular rodent diet and given drinking water ad libitum . These mice were used to collect primary peritoneal macrophages . All experiments were performed according to National/EU guidelines regarding the care and use of laboratory animals in research . Approval for these studies was obtained from the Ethics Committee of the Spanish National Research Council ( CSIC file CEA-213-1-11 ) . Mouse peritoneal macrophages were obtained as described previously [17] and plated at a density of 1×105 macrophages/well in RPMI 1640 medium supplemented with 10% HIFBS , 2 mM glutamate , penicillin ( 100 U/mL ) and streptomycin ( 100 µg/mL ) in 24-well tissue culture chamber slides . Late-stage promastigotes from WT and resistant lines were used to infect macrophages at a macrophage/parasite ratio of 1∶10 . Eight hours after infection at 35°C in an atmosphere containing 5% CO2 , extracellular parasites were removed by washing with serum-free medium . Infected macrophage cultures were maintained in RPMI 1640 medium plus 10% HIFBS at 37°C with 5% CO2 at different drug concentrations . After 72 h , macrophages were fixed for 30 min at 4°C with 2 . 5% paraformaldehyde in phosphate-buffered saline ( PBS; 1 . 2 mM KH2PO4 , 8 . 1 mM Na2HPO4 , 130 mM NaCl , and 2 . 6 mM KCl adjusted to pH 7 ) , and permeabilized with 0 . 1% Triton X-100 in PBS for 30 min . Intracellular parasites were detected by nuclear staining with Prolong Gold antifade reagent plus DAPI . Drug activity was determined from the percentage of infected cells and the number of amastigotes per cell in drug-treated versus non-treated cultures [17] . The levels of non-protein thiols were measured by flow cytometry using CellTracker , as described previously [18] . Parasites ( 107 promastigotes/mL ) , grown in M199 medium plus 10% HIFBS were washed twice with PBS and incubated with 2 µM CellTracker for 15 min at 37°C . They were then washed again with PBS and analysed by flow cytometry in a FACScan flow cytometer ( Becton-Dickinson , San Jose , CA ) equipped with an argon laser operating at 488 nm . Fluorescence emission between 515 and 545 nm was quantified using the Cell Quest software . Non-protein thiol-depleted parasites obtained after incubation with 3 mM BSO ( a γ-glutamylcysteine synthetase inhibitor ) for 48 h at 28°C were used as controls . ATP was measured using a CellTiter-Glo luminescence assay , which generates a luminescent signal proportional to the amount of ATP present , as described previously [19] . Promastigotes ( 4×106/mL ) were incubated at 28°C in RPMI plus 20% HIFBS containing 0 . 2 µM AmB or 25 µM MLF for 3 h , or 2 mM SbIII for 8 h . The drug concentration and incubation time were selected by monitoring parasite viability under a microscope . A 25-µL aliquot of parasites was then transferred to a 96-well plate , mixed with the same volume of CellTiter-Glo , and incubated in the dark for 10 min . The resulting bioluminescence was measured using an Infinite F200 microplate reader ( Tecan Austria GmbH , Austria ) . ΔΨm was measured by flow cytometry using Rh123 accumulation , as described previously [20] . The parasites ( 4×106 promastigotes/mL ) were incubated with the drugs as described above . 0 . 5 µM Rh123 was then added and the parasites incubated for a further 10 min . They were then washed twice , resuspended in PBS and analysed by flow cytometry in a FACScan flow cytometer ( Becton-Dickinson , San Jose , CA ) equipped with an argon laser operating at 488 nm . Fluorescence emission between 515 and 545 nm was quantified using the Cell Quest software . Parasites fully depolarized by incubation in 10 µM FCCP for 10 min at 28°C were used as controls . Statistical comparisons between groups were performed using Student's t-test . Differences were considered significant at a level of p<0 . 05 .
The resistant lines were selected in vitro in L . donovani promastigotes by a stepwise adaptation process , with drug concentrations starting below the EC50 values and gradually increasing , over 10 weeks ( equivalent to 90 generations ) , to a maximum concentration of 0 . 1 , 8 , 20 and 80 µM for AmB , MLF , PMM and SbIII , respectively . Resistance to single drugs and to double drug combinations was induced . The singly AmB-resistant line ( A ) and the lines resistant to the combination of AmB with MLF ( AM ) , PMM ( AP ) or SbIII ( AS ) showed similar EC50 values for AmB of 0 . 14 µM , a value two-fold higher than for the WT line ( Table 1 ) . In contrast , the singly MLF-resistant line ( M ) and the lines resistant to the combination of MLF with AmB ( AM ) or PMM ( MP ) showed EC50 values for MLF 1 . 81 , 3 . 10 , and 4 . 43-fold higher than for the WT line , respectively ( Table 1 ) . Likewise , the singly PMM-resistant line ( P ) and the lines resistant to combinations with AmB ( AP ) , MLF ( MP ) or SbIII ( SP ) , showed EC50 values for PMM 11 . 02 , 2 . 12 , 14 . 25 , and 18 . 35-fold higher than for the WT line , respectively ( Table 1 ) . Finally , the line resistant to SbIII alone ( S ) and the lines resistant to SbIII in combination with AmB ( AS ) or PMM ( SP ) showed EC50 values for SbIII that were 3 . 04 , 2 . 05 , and 2 . 18-fold higher than for the WT line , respectively ( Table 1 ) . All the resistant lines showed a similar growth rate , morphology , motility and macrophage infectivity to the WT line ( data not shown ) . We undertook additional experiments to determine whether the resistance to single drugs and drug combinations shown by the promastigote forms was maintained in intracellular amastigotes obtained after infection of mouse peritoneal macrophages ( Table 2 ) . The results indicated that the resistance indices to drugs in the different resistant intracellular amastigote lines were maintained and were very similar to those observed in their promastigote counterparts ( Tables 1 and 2 ) . The exception was the S-resistant line , which showed a significantly higher resistance index to SbIII ( 12-fold , Table 2 ) than that observed for the promastigote form ( 3-fold , Table 1 ) . The higher resistance index for PMM in the singly P-resistant line ( 11-fold ) and in the MP- ( 11- and 14-fold for intracellular amastigotes and promastigotes , respectively ) and SP- resistant lines ( 16- and 18-fold for intracellular amastigotes and promastigotes , respectively ) it worthy of note . Furthermore , a comparison of AP with P , and AS or SP with S , shows that the doubly resistant lines exhibit lower EC50 values for PMM or SbIII than their singly resistant counterparts . In contrast , the SP line exhibits a higher EC50 value for PMM than the P line ( Table 2 ) . The resistant phenotypes were stable in a drug-free medium for 1 month in the singly A-resistant and AM , AP , and AS doubly-resistant lines for AmB . In contrast , the remaining resistances were unstable , although the EC50 values were higher than for the WT line , except for the M and AP lines , which lost resistance against MLF and PMM , respectively ( Figure 1 ) . After culture for four months in a drug-free medium , all lines lost their resistance levels either completely or partially , except the AP line , which maintained a similar initial resistance level for AmB ( Figure 1 ) . These findings suggest that the resistance phenotype of the induced drug combination resistance is unstable . We investigated the cross-resistance profile of the promastigote and intracellular amastigote forms of each resistant line to different anti-leishmanial drugs ( Tables 1 and 2 ) . Both the A- and S-resistant lines showed a significant cross-resistance profile to PMM ( Table 1 ) , and the P-resistant line showed resistance to SbIII ( Table 1 ) . However , the M and P lines only showed resistance to SbIII in their intracellular amastigote forms ( Table 2 ) . In the case of drug combination resistant lines , we found that the AP-resistant line showed significant cross-resistance to MLF in both its promastigote and intracellular amastigote forms . Similarly , the AS-resistant promastigote line shows cross-resistance to PMM , and the AM-resistant promastigote line shows cross-resistance to PMM and SbIII ( Tables 1 and 2 ) . However , the MP and SP lines showed no cross-resistance to other anti-leishmanial drugs as either promastigotes or intracellular amastigotes ( Tables 1 and 2 ) . Our results concerning the stability of the cross-resistance in resistant lines maintained without drug pressure for one month showed that all lines maintained their resistance , except the AM line , which lost its cross-resistance to SbIII ( Figure 2 ) . An increase in thiol levels has been considered to be one of the main detoxification mechanisms observed in lines selected for resistance to SbIII [21] . In light of this , we determined the total intracellular non-protein thiol content in the different resistant promastigote lines using CellTracker . The results of this study showed significantly higher thiol levels in the resistant lines than in the WT line , except for the M and S lines ( Figure 3 ) . The highest thiol values were found in the MP and SP lines . As expected , a drastic decrease in thiol content was observed in all lines after incubation with BSO ( Figure 3 ) . We assessed the ATP levels in the presence of AmB , MLF and SbIII , which are known to induce an apoptotic-like process associated with ATP depletion in Leishmania [22] . PMM was not assessed as this drug kills the parasites by a different mechanism [22]–[24] . The WT parasites exhibited a significant decrease in total ATP levels after treatment with AmB , MLF or SbIII , with the former showing the highest decrease ( Figure 4A ) . The lines resistant to single and drug combinations remained protected or were more tolerant to ATP loss ( Figure 4A ) , with the AS-resistant line in particular showing a very small decrease in ATP levels after treatment with AmB or SbIII . A similarly small decrease in ATP levels was also observed after treatment of the M- and MP-resistant lines with MLF , and the S-resistant line with SbIII ( Figure 4A ) . Additionally , we observed that the M- and S-resistant lines presented significantly higher basal ATP levels without drug pressure ( Figure 4A ) , thus suggesting that these resistant lines have developed , amongst other resistance mechanisms , an increase in ATP levels . In contrast , the AM- and AP-resistant lines presented significantly lower basal ATP levels ( Figure 4A ) . ΔΨm is essential to mitochondrial ATP synthesis and changes to it are one of the markers for apoptosis induced by exposure to AmB , MLF and SbIII ( but not exposure to PMM ) [22] . Furthermore , mitochondrial oxidative phosphorylation in Leishmania accounts for most of the ATP expenditure of Leishmania parasites [25] . As a result , we tested the ΔΨm of WT and the various resistant lines by measuring Rh123 accumulation ( Figure 4B ) . WT parasites incubated with AmB , MLF or SbIII showed a significant decrease in Rh123 accumulation ( 7 . 7- , 2 . 6- and 1 . 4-fold , respectively ) . These values ( except for that for SbIII ) were even lower than those obtained upon incubation of parasites with the control uncoupling reagent FCCP ( Figure 4B ) . Except for the M-resistant line , the untreated resistant lines showed a significantly lower accumulation of Rh123; however , after treatment with the different anti-leishmanial drugs , the resistant lines showed a lower reduction ratio of Rh123 accumulation than the WT line ( Figure 4B ) . Consequently , the resistant parasites remain protected against the oxidative stress induced by treatment with the anti-leishmanial drugs AmB , MLF and SbIII .
Drug combinations for the treatment of leishmaniasis represent a promising and challenging chemotherapeutic strategy that has recently been implemented in different endemic areas . This approach has several advantages over single-drug therapies , including shortening of the treatment period and reduction of the probability of selecting drug-resistant parasites . However , this approach must be used with care given to the possibility that , if not applied in a controlled and regulated way , resistance could be induced in Leishmania , thus resulting in a rapid loss of efficacy of not one but two therapeutic options [13] . It is therefore important to design relevant experimental studies in order to determine whether Leishmania parasites are able to develop resistance to the different potential anti-leishmanial drug combinations that are likely to be used in the near future . The results obtained from such experimental studies could help to predict the likely success of drug combination therapy . There is still a great deal of debate concerning the clinical relevance of findings in promastigotes since this is not the stage that will eventually become exposed to the drug . It has recently been shown that differences can be obtained during the experimental induction of resistance to PMM using promastigotes and intracellular amastigotes depending on the resistance-selection protocol [26] . The methodology and technical difficulties required in inducing resistance to drug combinations justify the use of promastigote forms in this manuscript . However , experiments using intracellular amastigotes derived from resistant promastigotes could be useful when considering future recommendations for optimal drug combinations to combat different Leishmania species . Studies in clinically resistant isolates , where the mechanisms of resistance involve multi-factorial events that contribute to the tolerance to chemotherapeutic agents in Leishmania , are somewhat more complex [27] , [28] . In this paper , we have induced experimental resistance to the drug combinations AM , AP , AS , MP and SP in L . donovani ( MHOM/ET/67/HU3 , also known as LV9 or L82 ) . The sensitivity values of the parental L . donovani strain to the different anti-leishmanial drugs used were similar to , or even lower than , the published data for this and other L . donovani strains [11] , [26] , [29]–[31] . It is important to point out that this is the first description of an experimental induction of resistance to a combination of different anti-leishmanial drugs . Consequently , the conditions and times required for the induction of resistance can not be compared to the previously described induction of resistance to single anti-leishmanial drugs . Additionally , the single-drug resistance studies in L . donovani described by different groups , obtained higher levels of resistance after increasing the drug pressure and exposure times . Thus , an approximately 14-fold resistance has been obtained for MLF resistance [32] , an approximately 5- to11-fold resistance for PMM , depending on the L . donovani strain used [26] , and an approximately 20-fold resistance for AmB [31] . Moreover , it is important to note that , in clinical isolates resistant to sodium stibogluconate , an up to 41-fold higher tolerance to SbIII has been observed with respect to the susceptible clones of promastigote forms in the stationary growth phase [27] . In summary , the results of this study show that L . donovani can easily develop resistance to the drug combinations MLF/PMM and SbIII/PMM , with higher resistance indices than those found for AmB/MLF , AmB/PMM or AmB/SbIII . These results have been validated in intracellular amastigotes and are of considerable interest for future applications . Experimental resistance of L . donovani to the drug combination MLF/PMM , a combination that could , in theory , have advantages over other drug combinations as regards future use , is easily achieved . Similarly , the experimental studies described herein confirm how the ease with which experimental resistance to SbIII/PMM is induced in Leishmania . These studies should therefore be taken into account when it comes to future recommendations for their use in endemic areas , especially as the SbIII/PMM combination appears to be effective against VL ( in East Africa ) and has the additional advantage of low cost [7] . Consequently , further research into combination regimens when given for a short period and at lower total dose are required . We have also confirmed that Leishmania-resistant parasites develop an increase in cellular thiol redox metabolism as a drug-detoxification mechanism to protect against drug-induced loss of ATP and mitochondrial membrane potential . As described previously , the anti-leishmanial drugs AmB , MLF and SbIII ( but not PMM ) induce a significant decrease in the mitochondrial membrane potential , thus leading to a bioenergetic collapse of the parasite and drug-induced cell-death [22] . A link between the mode of killing of drugs against Leishmania infantum ( such as AmB , MLF and Sb , which share a similar mode of killing ) , and the tolerance towards cell death induced by their respective anti-leishmanial drugs , has been described previously [22] . Although , in contrast to PMM , this was thought to facilitate the emergence of multidrug resistance , similar findings were not observed under our experimental conditions . Instead , our results show a significant cross-resistance profile to PMM in the AM- and AS-resistant lines and to MLF in the AP-resistant line . Conversely , the AP-resistant intracellular amastigote line acquired cross-resistance to MLF . A similar absence of a link between cross-resistance to drugs with similar mechanism-of-death pathways was observed in the singly A- and S-resistant promastigote lines , with a cross-resistance to PMM , and in the P-resistant line , with resistance to SbIII . The absence of any such correlation could be explained by taking into account that different Leishmania species present different drug susceptibilities and different abilities to respond to drug pressure . In light of the characteristics of this infectious disease and the existence of different Leishmania species , with their different drug susceptibilities , it is possible that each Leishmania species will require a different drug combination . Suitable options for combination treatment must therefore be optimised in further experimental studies . In this respect , genome-sequencing and metabolomics experiments are currently underway to determine the specific resistance mechanisms developed by Leishmania parasites to different drug combinations . Finally , in view of the proven value of these results for the research community , and considering the debate as regards the use of promastigotes or intracellular amastigotes for induction of drug resistance to drug combinations , we are currently attempting to induce resistance to drug combinations in intracellular amastigotes , as the results obtained will be of greater significance in terms of the conditions found in the field . | Leishmania is a protozoan parasite that infects human macrophages to produce the neglected tropical disease known as leishmaniasis . Chemotherapy is currently the only treatment option for leishmaniasis . First-line therapies include pentavalent antimonials , except in some regions in the Indian subcontinent , the liposomal formulation of amphotericin B , miltefosine and paromomycin . The WHO has recently recommended a combined therapy in order to extend the life expectancy of these compounds . However , resistance could be induced in Leishmania if this approach is not applied in a controlled and regulated way , thus resulting in a rapid loss of efficacy of not one but two therapeutic options . In light of this , we have designed relevant experimental studies in order to determine whether Leishmania parasites are able to develop resistance to the different potential anti-leishmanial drug combinations that will be used in the near future . The results obtained could help us to predict the success of drug combination therapy . Experimental resistance of Leishmania donovani promastigotes to drug combinations was obtained after 10 weeks and remained in the intracellular amastigotes . We therefore conclude that L . donovani can easily develop resistance to drug combinations mainly miltefosine/paromomycin and SbIII/paromomycin . These results have been validated in intracellular amastigotes and are of considerable interest for future prediction of the success of drug combination therapy . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"leishmaniasis",
"neglected",
"tropical",
"diseases",
"parasitic",
"diseases"
] | 2012 | Leishmania donovani Develops Resistance to Drug Combinations |
A deterministic population dynamics model involving birth and death for a two-species system , comprising a wild-type and more resistant species competing via logistic growth , is subjected to two distinct stress environments designed to mimic those that would typically be induced by temporal variation in the concentration of a drug ( antibiotic or chemotherapeutic ) as it permeates through the population and is progressively degraded . Different treatment regimes , involving single or periodical doses , are evaluated in terms of the minimal population size ( a measure of the extinction probability ) , and the population composition ( a measure of the selection pressure for resistance or tolerance during the treatment ) . We show that there exist timescales over which the low-stress regime is as effective as the high-stress regime , due to the competition between the two species . For multiple periodic treatments , competition can ensure that the minimal population size is attained during the first pulse when the high-stress regime is short , which implies that a single short pulse can be more effective than a more protracted regime . Our results suggest that when the duration of the high-stress environment is restricted , a treatment with one or multiple shorter pulses can produce better outcomes than a single long treatment . If ecological competition is to be exploited for treatments , it is crucial to determine these timescales , and estimate for the minimal population threshold that suffices for extinction . These parameters can be quantified by experiment .
To quantify pulse efficiency , we primarily study the minimal population size nmin of our two-species system as a proxy for the extinction probability of the population . Antibiotic stewardship programmes suggest that for some diseases , such as pneumonia , the immune system can clear the residual infection once the bacterial population size is sufficiently reduced [19 , 20] . Thus , the minimal population size may be a more relevant parameter than the exact extinction probability itself . Additionally , the general behaviour of the deterministic system and its observable nmin is more robust than the extinction probability in the stochastic model , as in the latter , the precise form of stochastic noise , or the system size , would be important . The total population minimum nmin can still serve to gauge the latter , which scales as exp ( nmin ) [21] . Before introducing our model in detail , we jump ahead and summarise the essential result of our work in Fig 1 , which we discuss later in more detail . Fig 1 shows the value of the minimal total population size in the configuration space of drug concentration profiles , spanned by the width of the pulse ( or duration of its high stress environment ) on the x-axis and the form of the pulse on the y-axis , which will be explained later . Practically , these two properties of a pulse—its high stress duration and its form—are likely constrained: a very long duration of the high stress environment or stronger drug might be detrimental for patients due to , for example , a destructive impact on the gut microbiome [22 , 23] . Similarly , some pulse forms , such as those where the highest possible drug concentration suddenly drops to zero at the pathogen location at the end of the pulse ( here denoted by temporal skewness s = 1 ) , may not be realistic for clinical treatments . However , since we do not want to make any assumptions on which parts of the configuration space should be accessible , we examine our system for all possible combinations of pulse form and durations of the high stress environments . The colour code ( symbols ) signifies which of the four possible pulse sequences sketched on the right of Fig 1 most effectively reduces the population size . Fig 1 clearly shows that in our simple model setup , different pulse sequences are favourable in different regions of configuration space . The aim of this work is to outline phenomenologically which pulse sequence yields the lowest minimal population for which part of configuration space in Fig 1 , and might therefore be most likely to drive the species to extinction . The best pulse sequence at any one point of Fig 1 tends to be the one that maximally exploits the competition between the more resistant and wild-type species , represented by logistic growth in our model . The simplicity of our approach makes explicit why some references might argue for more moderate treatments involving e . g . shorter or lower drug concentrations , but also what the limitation of models and observables are and hence why such moderate treatments may not work in real setups . We also examine how the population composition ( a measure of how strongly the more resistant species dominates the population ) evolves , should such a pulse sequence not lead to achieving extinction . Finally , we highlight the need for microbial experiments in such temporally varying drug gradients , in order to evaluate the applicability of simple models to real systems .
The simplest model that can be used to study the effect of the temporal concentration profile on a heterogeneous population ( n ) consists of two phenotypically different species , a susceptible “wild type” species ( w ) , and a more tolerant or resistant species ( r ) . Its increased resistance comes at the cost of a reduced fitness in the drug-free environment , which is reflected in a smaller growth rate . As in previous works [24 , 25] , we assume that the drug is bacteriostatic , that is , it only affects growth , such that growth of each species ceases as soon as its minimum inhibitory concentration ( MIC ) is exceeded . Thus , in this deterministic population dynamics model for the birth-death process , sketched in the inset of Fig 2 , the growth rate of each species η ∈ {r , w} is given by ϕη ( t , n ( t ) ) = Θ[MICη − c ( t ) ]λη ( 1 − n ( t ) ) , where n ( t ) = w ( t ) + r ( t ) is the total number of species at time t expressed in terms of a carrying capacity , which does not require specification as it serves merely as a unit for the population size . The Heaviside-Theta step function Θ implies that the growth rate is only non-zero when the drug concentration is lower than the MIC of the corresponding species . The index η ∈ {r , w} refers to the type of species ( resistant or wild-type ) , and λη is its growth rate . The more resistant species has a lower basal growth rate in the drug-free environment , i . e . , λr = λw − k≔ λ − k , where k > 0 can be interpreted as a cost that the resistant species incurs for being more resistant . The logistic growth assumed in this model introduces competition between the wild-type and the resistant species for limited space and/or resources , and places an upper bound on the population size . We also include a constant death rate δ for both species , meaning that a species decays at rate δ when c ( t ) >MICη: For these higher concentrations , growth of species η is inhibited and , since switching is negligible , the species can only die . All rates and times in this work are given in units of λ . The time evolution of the population can be studied in terms of the differential equations w ˙ ( t ) = [ ϕ w ( t , n ( t ) ) - δ - μ w ] w ( t ) + μ r r ( t ) r ˙ ( t ) = μ w w ( t ) + [ ϕ r ( t , n ( t ) ) - δ - μ r ] r ( t ) ( 1 ) since for sufficiently large populations stochastic fluctuations can be neglected . The two species are coupled via the competition from logistic growth , as well as via the switching rates μw and μr . Phenotypically more resistant states can be characterised by a reduced growth rate , or complete growth arrest , often known as tolerance or persistence [26–28] ( for a recent review , see Ref . [29] ) . Provided that μw , r ≪ δ , which is the case for both mutation and phenotypic switching , our choice of μw = 10−6 λ and μr = 0 does not qualitatively affect the results . For this entire work , we used exemplary values of δ = 0 . 1λ and k = 0 . 1λ , where λ ≡ 1 , i . e . we used λ as the basic unit of time . We investigate several other combinations of costs and death rates , in particular combinations with the same death rate , but a smaller and larger cost , in S1 Text . There , we show that our results and general statements are still valid for these cases . We chose the values of δ = 0 . 1 and k = 0 . 1 since this combination allowed us to show the complete and most general picture of possible best pulse shapes in Fig 1 . A smaller ( yet also biologically possible ) fitness cost would not have contained all different scenarios . We ask the reader to refer to S1 Text for more details . Since in our model the only relevant information about the antibiotic concentration is whether it is above or below the MIC of the corresponding species , any pulse sequence is fully determined by the temporal arrangement of low-stress ( low ) and high-stress ( high ) environments . In these ( low ) and ( high ) environments , the antibiotic concentration is low , MICw < c ( t ) < MICr , or high , c ( t ) > MICr , respectively ( sketched for a single pulse in the top panel of Fig 2 ) . Before the pulse sequence , the system is in the drug-free environment ( free ) , where the concentration of the antibiotic c ( t ) is less than either MIC , c ( t ) < MICw , r . We assume that the ( free ) environment appears only before , but not during , a pulse sequence . Thus , the ( free ) environment determines the initial condition of the population , which we take to be at its fixed point , ( w ( t = 0 ) , r ( t = 0 ) ) = ( w ( free ) * , r ( free ) * ) , shown as the purple dot near the w-axis of the phase space panel ( free ) of Fig 2 . The change in population size and composition in each of these environments is characterised by the flow field in phase space ( w , r ) , shown in the three lower panels of Fig 2 . In the ( low ) environment , the population flows towards the more resistant species ( high r and low w ) , while in the ( high ) environment , it flows towards the origin , meaning that both species die out exponentially . Thus , the effect of a single pulse on the population crucially depends on the times spent by the system in the ( low ) and ( high ) environments . A single pulse involves a single ( connected ) environment of ( high ) antibiotic exposure , with ( low ) environment potentially preceding or succeeding this ( high ) environment . In reality , the duration of these ( low ) environments will depend on the experimental setup or host . A pulse sequence is composed of a succession of identical single pulses . We refer to the total time of the pulse as τ , and the time during which the system is in the ( high ) environment as tr . The time periods during which the system is in the ( low ) environment ( initially ) before tr and ( finally ) after tr are denoted by t w ( i ) and t w ( f ) , respectively . As this would overparameterise the pulse , we combine the latter two time scales into a skewness parameter s = ( t w ( i ) - t w ( f ) ) / ( τ - t r ) , signifying how tr is positioned within τ . Skewness s = − 1 ( s = 1 ) thus denotes a pulse which starts ( ends ) with the ( high ) environment , while skewness s = 0 denotes a symmetric pulse .
We compared pulse sequences of up to N = 4 pulses ( same tr and s ) for constant treatment time τ for all possible skewnesses s and durations tr . Thus , a single pulse with τ = 60 and given tr and s is compared with a sequence of N identical pulses , each defined by τ ( N ) = 60/N and t r ( N ) = t r / N and s . ( Further values of τ are discussed in S1 Text ) . The retention of the same skewness within a sequence is motivated by the fact that we assume that the rate of increase or decrease in concentration is primarily determined by the host system of the bacteria . In this comparison , the ‘best’ pulse sequence for given ( tr , s ) is defined as the one that yields the lowest population minimum nmin and so has the highest likelihood of eliminating the pathogen . In situations where the entire configuration space is accessible , the maximal tr yields the overall lowest population minimum , independent of skewness s . Since practically the maximal duration tr acceptable for treatments may be limited , it is important to know which pulse sequence is best for each ( tr , s ) , such that we can provide intuition on any situation and parameter choice that may arise . The colour ( and corresponding symbols ) in Fig 1 show the best pulse sequences ( i . e . the best N ) , and the shade indicates the value of nmin ( dark denotes high values ) . We found that a single pulse is most effective over a large range of parameters ( blue in Fig 1 ) . In particular , for each duration in the ( high ) environment tr , the lowest minimum across all skewnesses is obtained by a single pulse ( blue line ) . This means that in practical situations which allow all different pulse skewnesses , a single pulse with a skewness on the blue line would give the lowest minimum . If , however , the possible pulse skewness is limited due to the host setup , a single pulse may not be the best choice . For ease of comparison , Fig 3 a shows nmin for just a single pulse of constant treatment time τ = 60 , with the white line marking the lowest minimum ( the blue line in Fig 1 ) . In the next paragraph we focus on a single pulse in order to understand which pulse parameters ( s , tr ) yield this lowest minimum . In the previous section , we learnt which pulse sequences yield the maximal relative reduction in population size for which regions in ( tr , s ) -space . This minimal population nmin served as a proxy for gauging when extinction would most likely occur in a setting where an immune response can destroy the population when it is already small . Now , we would like to address a complementary question: in the event that extinction does not occur , whether because nmin was too high , or the population was small for too short , what is the effect of such a ‘failed’ pulse on the bacterial population ? We already saw that the composition of the population shifts more towards r with each pulse . In terms of real treatments , it might often be better not to pursue treatments which , if unsuccessful , entail a high risk of creating a fully resistant population . In order to evaluate the pulsed treatments associated with the most effective population reductions based on Fig 1 , we now focus on the population composition , quantified by the ratio of resistant to wild-type species , r/w , at the end of the best pulse within the best sequence . Evaluating r/w at the end of the pulse that yields the global minimum is motivated by the fact that the treatment can be stopped after , but not during , an individual pulse . Fig 5b shows the dependence of r/w on the pulse configuration , which can be best understood by first considering how the population evolves in ( w , r ) phase space during the different pulse sequences . In Fig 5a , we show trajectories for three pulse sequences , consisting of a single , two , or three pulses respectively , with τ = 60 and tr = 10 . The qualitative behaviour of the phase space trajectory is independent of skewness ( in Fig 5a , s = 0 . 9 , the corresponding trajectories for s = −0 . 5 , s = 0 . 2 and s = 0 . 5 can be found in Fig D in S1 Text ) . The colour of the trajectory darkens progressively with every pulse in the sequence . The trajectory starts at the ( free ) fixed point close to the w-axis beyond the limits of Fig 5a , and evolves towards the r-axis . Within each sequence , r/w steadily increases from pulse to pulse , as r progressively takes over the population during the ( low ) regimes . Thus , in the top left corner of Fig 5b , where the first pulse of the sequence yields the lowest minimum , r/w is comparatively smaller ( lighter shading ) . Indeed , the higher the N of the best sequence , the lower the ratio in Fig 5b , provided the global minimum is reached in the first pulse ( such as in the red region in Fig 1 ) . The region marked with the white line in Fig 1 , where intermediate pulses ( and not the first pulse ) in the sequence yielded the lowest global minimum , also shows up clearly as darker in Fig 5b . Here , r has grown more than for a single pulse , as more pulses were applied before the population minimum was reached . Thus , in our model , when both population reduction and composition are considered , pulse sequences where the minimum is attained in the first pulse are generally more effective than a single long pulse: maintaining the ( low ) regime in the first pulse for around to keeps r/w as well as nmin small . This argument suggests that treating with this first pulse only achieves the best result , and additionally comes with a shorter total treatment duration τ and a shorter tr . We would like to note that even if the population does not die out during this short treatment , multiple pulses of this form could be added in order to give the immune system more opportunities to eliminate the infection . These additional pulses would not drastically change r/w compared to the composition obtained after the single long pulse of τ = 60 . This can be seen also in Fig 5a , where for all pulse sequences the population composition is similar at the end of the entire sequence .
Experiments with microbes can help investigate minimal antibiotic dosages and treatment times in a well-controlled test tube setup , where the impact of certain treatments on the microbial species itself can be studied without interfering effects , for example from the immune system . Such microbial experiments have , for example , helped suggest drug combinations or treatment regimens which could retard the development of antibiotic resistance [30–33] . Increasingly , these experiments try to incorporate practically important aspects of heterogeneities in the environment [34] , such as drug concentration gradients . These gradients can enhance the development of bacterial resistance relative to spatially homogeneous systems [24 , 25 , 35 , 36] , as the more resistant species can successfully compete with a faster growing , but more susceptible wild-type species . Not enhancing the selective advantage of the more resistant species , in the context of temporal heterogeneities in drug levels , including the duration , frequency and even the concentration profile during a single antibiotic pulse , as studied also in this work , is also important in real treatments [7 , 37] , and is thus within the limits of current experiments . Our model makes two drastic simplifications compared to real microbial species . First , we study only two species , instead of a series of possible phenotypically or genotypically different species . Typically , the evolutionary pathway that leads to a fully resistant species involves a variety of intermediate mutants , even when the mutational paths are constrained [38] . Since the fitness benefit diminishes with each successive mutation in a series [39 , 40] , we assumed that the strongest effect is conferred by the first mutation , and neglected all higher order mutants . For phenotypic switches , it is reasonable to consider only two species , corresponding to , for example , the expression or repression of a protein [41 , 42] . Thus , our model should be applicable to experimental systems , while in real patients , different types of tolerant or persister cells might be involved [43] , or even interact [44] . The second simplification concerns how these two species are affected by the antibiotic . In our model , we assume that the antibiotic is bacteriostatic , i . e . only affects the growth of the species [24 , 25] . We also assume that the growth rate of each species falls abruptly to zero when the antibiotic concentration is higher than their respective minimum inhibitory concentration ( MIC ) ( see e . g . [45] ) . The experimental situation is more complex: cessation of growth is not instantaneous , the space occupied by a dead cell may not immediately become available [42] , and the general use of the MIC as an indicator for slow growth is questionable [46] . However , an abrupt change in growth rate at MIC has been verified experimentally for E . coli and chloramphenicol . [25] . Additionally , our analysis is based on large numbers rather than extinction events which would be model specific; thus , small changes in the model ( such as reduced but non-zero growth rate ) should still give qualitatively similar results . Evaluating the effect of different pulse sequences should be possible within a microfluidics setup , where , for example , periodically fluctuating environments have already been investigated for E . coli and tetracycline [42] . We expect that one should be able to observe that the ( low ) environment of drug concentration can be exploited in order to increase extinction probabilty for a ( high ) environment that is present for as short as possible , with the treatment time being constant . How long this duration of the ( low ) environment is for best exploitation would be sensitive to the growth rate of the more resistant species , which for tetracycline could be generated using a specific promoter , namely the agn43 promotor [42 , 47 , 48] . Just as shown in Fig 1 , we expect higher N pulse sequences to do better when this duration is optimal for them , but not for the longer pulse . In addition , further study of E . coli in combination with other antibiotics and more resistant strains should also show this , in addition to being more realistic than our simple model . | The possibilities of lower antibiotic dosages and treatment times , as demanded by antibiotic stewardship programmes have been investigated with complex mathematical models to account for , for example , the presence of an immune host . At the same time , microbial experiments are getting better at mimicking real setups , such as those where the drug gradually permeates in and out of the region with the infectious population . Our work systematically discusses an extremely simple and thus conceptually easy model for an infectious two species system ( one wild-type and one more resistant population ) , interacting via logistic growth , subject to low and high stress environments . In this model , well-defined timescales exist during which the low stress environment is as efficient in reducing the population as the high stress environment . We explain which temporal patterns of low and high stress , corresponding to sequences of drug treatments , lead to the best population reduction for a variety of durations of high stress within a constant long low stress environment . The complexity of the spectrum of best treatments merits further experimental investigation , which could help clarify the relevant timescales . This could then give useful feedback towards the more complex models of the medical community . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"antimicrobials",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"drugs",
"immunology",
"microbiology",
"antibiotic",
"resistance",
"probability",
"distribution",
"mathematics",
"pharmaceutics",
"antibiotics",
"pharmacology",
"populati... | 2017 | Exploiting ecology in drug pulse sequences in favour of population reduction |
Transfer RNAs ( tRNAs ) are ancient molecules that are central to translation . Since they probably carry evolutionary signatures that were left behind when the living world diversified , we reconstructed phylogenies directly from the sequence and structure of tRNA using well-established phylogenetic methods . The trees placed tRNAs with long variable arms charging Sec , Tyr , Ser , and Leu consistently at the base of the rooted phylogenies , but failed to reveal groupings that would indicate clear evolutionary links to organismal origin or molecular functions . In order to uncover evolutionary patterns in the trees , we forced tRNAs into monophyletic groups using constraint analyses to generate timelines of organismal diversification and test competing evolutionary hypotheses . Remarkably , organismal timelines showed Archaea was the most ancestral superkingdom , followed by viruses , then superkingdoms Eukarya and Bacteria , in that order , supporting conclusions from recent phylogenomic studies of protein architecture . Strikingly , constraint analyses showed that the origin of viruses was not only ancient , but was linked to Archaea . Our findings have important implications . They support the notion that the archaeal lineage was very ancient , resulted in the first organismal divide , and predated diversification of tRNA function and specificity . Results are also consistent with the concept that viruses contributed to the development of the DNA replication machinery during the early diversification of the living world .
Transfer RNA ( tRNA ) molecules are central to the entire translation process . They interact with the ribosomal RNA ( rRNA ) subunits as they are being ratcheted through the center of the ribosome [1] , [2] . Their acceptor arms charge specific amino acids through the activity of cognate aminoacyl-tRNA synthetases , while triplets of bases on their ‘anticodon’ arms recognize complementary ‘codon’ sequences in messenger RNA . These and many other molecular interactions define the identities and functions of these tRNA adaptors and establish a genetic code that translates nucleic acid into protein information in the cell . The structural make-up of tRNA is therefore fundamental to our understanding of how the biosynthetic machinery was set up into place in an emerging protein and organismal world . tRNAs are clearly ancient molecules [3] and they have been used profusely to study the evolution of ancient life [4]–[8] . The identity and function of tRNAs are fundamentally delimited by the structure of the molecules , and structure is more conserved than sequence . In fact , we recently showed that tRNA structure carries deep phylogenetic signal and can be used directly to reconstruct evolutionary history [9] . However , understanding phylogenetic trees is challenging because tRNA evolution embeds a history of recruitment in which structures gain or co-opt new identities and functions or takeover established ones . The hierarchical branching patterns of the universal tree of life portray the natural history of the living world . The current accepted universal tree proposes a tripartite world ruled by three superkingdoms , Archaea , Bacteria , and Eukarya [10] . This view stems fundamentally from the study of the small subunit of rRNA , a molecule that is also ancient and central to translation . The rise of evolutionary genomics with an analysis of entire repertoires of nucleic acid and protein molecules supports this tripartite scheme [11] , [12] . However , the root of the universal tree remains controversial and so is the nature of the universal ancestor of all life that this root defines [13] , [14] . We recently embarked on a systematic and global study of evolution of domain structure and organization in proteins [15] , [16] ( Wang and Caetano-Anollés , submitted ) . Structures were assigned to protein sequences in hundreds of completely sequenced genomes and a structural census of protein domains used to generate phylogenomic trees of protein architectures . The evolutionary genomic analysis defined a universal ancestor that was eukaryotic-like and had a relatively complex proteome [16] . It also established that the archaeal lineage was the most ancient and originated from reductive evolutionary tendencies in the use of protein architectures . In order to explore if similar phylogenetic signatures were present in tRNA , we apply a well-established cladistic method [17] , [18] that embeds structure directly into phylogenetic analysis [19] . The method involves identifying features characteristic of the secondary structure of RNA molecules , coding these features as linearly ordered multi-state characters , and using them to build phylogenetic trees with optimal tree search methods . The strategy has been used to reconstruct a tripartite tree of life from rRNA structure [17] , trace evolution of rRNA in ribosomes [18] , study the evolution of closely related phytopathogenic fungi [17] or distantly related members of the grass family [20] , and explore the structural origin and evolution of retrotransposons in eukaryotes [21] . We also used the approach to study the evolution of the major structural and functional components of tRNA , establishing that tRNA molecules originated in the acceptor arm and providing further support to the ‘genomic tag’ hypothesis [9] . Here we reconstruct global phylogenetic trees using information embedded in both the sequence and structure of tRNA molecules . As we have shown previously ( Sun and Caetano-Anollés , submitted ) , the intrinsically rooted trees revealed that tRNA with long variable arms ( known as class II or type II tRNA ) coding for amino acids Sec , Ser , Tyr , and Leu were ancient . However , trees failed to show clear patterns related to tRNA function , an observation that underscores the importance of recruitment and phylogenetic constraint ( factors that restrict the acquisition of phenotypic traits or functions in lineages ) in tRNA evolution . In order to sort out these confounding processes we built trees while forcing monophyletic groupings of taxa ( sets that share a common ancestor ) to test alterative hypotheses or establish evolutionary timelines of structural , functional , or organismal diversification . This strategy ( known as constraint analysis in phylogenetics ) provided an unanticipated window into early evolution of life .
Phylogenetic analyses of the combined dataset of sequence and structure of 571 tRNAs produced most parsimonious trees that were 10 , 083 steps in length and were intrinsically rooted ( Figure 1 ) . The tRNA set was obtained from Part 2 of the Bayreuth tRNA Database and represented organisms in the three superkingdoms of life and viruses and covered all isoacceptor variants and amino acid specificities ( Table S1 ) . This molecular set is unique since it contains information of modified bases and structures derived by comparative analysis ( see Materials and Methods ) . Bootstrap support ( BS ) values were generally low for most clades ( <50% ) , but this was generally expected given the large number of taxa ( molecules ) analyzed . Class II tRNA molecules with long variable arms , including tRNASec and most tRNASer , tRNATyr , and tRNALeu isoacceptors , appeared at the base of the rooted trees ( Figure 1 ) . Besides this pattern , trees failed to reveal groupings that would indicate clear evolutionary links to organismal origin or molecular functions . The monophyly of tRNA belonging to each superkingdom ( or viruses ) or expressing different amino acid specificities was not revealed . Similarly , tRNAs with specificities for amino acids defined previously as being ancestral [22]–[27] did not form monophyletic groups . tRNA molecules sharing the first , second , or first and second bases in codons were not monophyletic either . These patterns were also observed in trees derived from partitioned matrices of superkingdoms or viruses ( data not shown ) . In order to uncover deep phylogenetic signals and test alternative evolutionary hypotheses we forced groups of tRNAs that shared a same organismal origin ( molecules from each superkingdom of life or viruses ) into monophyly using constraint analyses . We then recorded the length of the most parsimonious trees that were obtained and the number of additional steps ( S ) that were needed to force the constraint . This exercise was generally done with or without forcing class I and II tRNA molecules into separate groups , but overall results were congruent . Constraints related to the diversification of the organismal world ( Table 1 ) consistently showed Archaea as the ancestral group ( i . e . , forcing archaeal tRNAs into monophyly was always associated with low S ) , followed by viruses , Eukarya , and Bacteria ( with S increasing in that order ) ( Figure 2 ) . Hypotheses of relationship among superkingdoms clarified further the possible rooting of the universal tree . Constraining molecules from Eukarya and Bacteria into a monophyletic group [constraint ( EB ) ] was the most parsimonious solution and was consistent with an early split of two ancient lineages , one of archaeal origin and the other of eukaryal-bacterial origin . When forcing molecules from two of the three superkingdoms individually and as a group into monophyly , the Eukarya and Archaea dichotomy [constraint ( ( E ) ( A ) ) ] was most parsimonious . This suggests the earliest two superkingdoms to diversify were Eukarya and Archaea . The S values for these constraints indicated that their diversification always preceded the onset of Bacteria . Finally , constraining molecules from the three superkingdoms into three separate groups in all possible 3-taxon statements showed that a polytomous arrangement was the most parsimonious . S values exceeded those indicating the onset of Bacteria as a group . These patterns maintained when tRNA structural categories were constrained and all phylogenetic statements were congruent ( Figure 2 ) . We also explored the origins of viruses by constraining molecules from each individual superkingdom or viruses into monophyletic groups , together [e . g . , ( AV ) ] or separately [e . g . , ( ( A ) ( V ) ) ] ( Table 2 ) . The most parsimonious scenario always linked the origins of viruses to the archaeal lineage , with S values matching those defining the organismal timeline ( Figure 2 ) . Dividing the viral sequences into two groups ( i . e . , viruses infecting Eukarya or Bacteria ) maintained the linkage between the separated groups of viruses and Archaea for various competing hypotheses ( Table 2 ) . Finally , we constrained trees according to isoacceptor group and then according to organismal group , or vice versa , with or without constraining tRNA categories ( Table 3 ) . A scenario in which organismal ( superkingdom ) diversification predated tRNA functional divergence was always more parsimonious ( S = 2 , 338–2 , 481 ) than one where functional divergence predated organismal diversification ( S = 2 , 415–2 , 534 ) . Since constraint analyses could be biased by unequal rates of evolution , we calculated average number of character change per branch in consensus trees generated from partitioned data matrices ( Table 4 ) . An analysis of variance ( ANOVA ) showed values were not significantly different in the three superkindoms of life and viruses ( p>0 . 05 ) . Similarly , we did not find differences when random trees were compared ( not shown ) .
In order to uncover evolutionary patterns related to organismal diversification , we first generated rooted phylogenetic trees using information embedded in the structure and sequence of tRNA ( Figure 1 ) . As expected , class II tRNA molecules with long variable arms coding for Sec , Ser , Tyr , and Leu appeared at the base of the rooted trees and were ancient . We also observed a rather tight paraphyletic clustering of tRNAs of archaeal origin . However , we were unable to reveal any other pattern of significance in the trees; no monophyletic groupings could be established when tracing tRNA function , codon identity , or organismal origin ( data not shown ) . In order to untangle the intricate history of tRNA , we forced trees to acquire pre-defined tree topologies representing competing ( alternative ) or non-competing phylogenetic hypotheses , constrained the exploration of tree space during phylogenetic searches , and produced sub-optimal tree reconstructions . Competing hypotheses were contrasted and those that imposed a minimum number of additional steps ( S ) on the optimal tree ( i . e . , more parsimonious ) were not rejected . Using this approach , we tested for example competing chronologies or sister taxa relationships related to organismal diversification . In turn , non-competing hypotheses were ranked by the values of S according to some external evolutionary model . In this study , they were used to define timelines of first appearance of superkingdoms and viruses in evolution . Hypotheses of origin that were satisfied with fewer steps were considered less affected by the confounding effects of recruitment in lineages and more ancient than those that required more steps . In other words , it was easy to merge lineages in backwards time ( a process known as coalescence ) to fit the constraint . Plots mapping the correlation between S and number of nodes from a hypothetical tRNA ancestor in the trees confirmed the validity of this assumption of ‘polarization’ ( data not shown ) . This type of analysis is not new . In cybernetics it is known as ‘constraint analysis’ and represents a formal method of decomposing a reconstructable system into its components by imposing natural or man-made limitations [28] . The method is widely used in cladistic and phylogenetic analyses to test for example hypotheses of monophyly [29] , but to our knowledge , has never been used to dissect systematically patterns in a phylogenetic tree . Two fundamental assumptions support the analysis . First , we assume tRNA structures acquired new identities and functions as the genetic code expanded , and that different structures were co-opted for the task in different lineages and different functional contexts . This assumption seems reasonable . Recruitment processes are common in evolution of macromolecules . In cellular metabolism , for example , enzymes are often recruited into different pathways to perform new enzymatic functions [16] , [30] , [31] . Moreover , structural diversification of tRNA appeared to have predated organismal diversification [32] ( Sun and Caetano-Anollés , submitted ) and the functions and identities attached to present-day tRNA structures probably developed in lineages and were shuffled by horizontal gene transfer . Second , we assume old tRNA structures developed or recruited new functions ( co-options ) more often than new tRNA structures acquired old functions ( takeovers ) . This assumption is also reasonable and appears to be supported by our studies of enzyme recruitment in metabolism ( Kim et al . , ms . in preparation ) . Our trees show several instances of takeovers , in which modern class I structures lacking the long variable arms took over ancient amino acid charging functions associated with class II structures ( Figure 1; Sun and Caetano-Anollés , submitted ) . However , old structures have more chances to succeed in a diversifying world , as they spread through lineages . Younger structures in turn are restricted to the lineage in which they originated , and can only spread further through horizontal transfer events . One implication of this assumption is that older functions will be less prone to co-options than younger functions , and that tRNA belonging to older lineages will be less affected by co-options than those in younger lineages . Consequently , ancient molecules sharing functions or belonging to selected lineages will be more easily constrained than younger variants in phylogenetic reconstruction . We also assume phylogenies are free from systematic errors and the confounding effects of mutational saturation , long branch attraction artifacts , and unequal rates of evolution along branches of the trees [11] . However , most branching events in these phylogenies happened a relatively long time ago and phylogenetic analyses of ancient molecules carry all the problems of deep reconstruction [33] . While the impact of some of these effects diminishes when using multi-state characters in tRNA structure [34] , [35] , different rates of change could alter the coalescense of lineages and our results . For example , increased rates of change known to occur in rapidly evolving viral molecules could increase expected S values , making the viral lineage artificially younger . Nevertheless , an analysis of rates of change in consensus and random trees derived from partitioned data matrices showed that evolutionary rates of tRNAs in the three superkingdoms of life or viruses were not significantly different in our analysis ( Table 4 ) . The fact that evolutionary rates in the four lineages were similar decreases the impact of unequal rates of evolution and underscores the conserved nature of tRNA structure when compared to sequence . Similarly , problems of statistical consistency related to long branch attraction could bias the reconstruction of the tRNA tree . These artifacts , which are rather common in sequence analysis , result from unequal rates of variation in branches and the interplay of short and long branches in a tree [36] . They are however not so much related to branch length ( which in our analyses do not vary considerably; Table 4 ) but to changes of a same character state occurring preferentially in long branches , forcing the tree-building method to join them artificially . However , the probability of these covarying homoplasies is known to decrease with increases in character states , as with the multi-state structural characters of this study [34] , and when branches are separated by increased taxon sampling [37]–[39] . Consequently , large trees as the tRNA trees we have reconstructed from sequence and structure in this study should be considerably less prone to consistency problems [38] , [39] than the four-taxon statements related to sequences originally used to define them [36] , especially if they involve multiple character states depicting structure . We constrained tRNA groups according to organismal origin using different schemes and tested possible competing and non-competing hypotheses describing timelines of organismal diversification and possible topologies of the universal tree of life ( Figure 2 ) . Constraining tRNAs belonging to individual superkingdoms or viruses showed Archaea as the most ancestral group , followed by viruses , Eukarya , and Bacteria , in that order . This timeline already suggests a very early split of the archaeal lineage in evolution . An analysis of the three possible two-superkingdoms single-group constraints showed that forcing molecules from Eukarya and Bacteria into a single monophyletic group [constraint ( EB ) ] was most parsimonious and confirmed the early split of lineages and separation of Archaea . It also suggested an important lineage relationship between Eukarya and Bacteria and a relative time frame for their coalescence as a group . Interestingly , S values for the eukaryal-bacterial lineage always coincided with those for the viral group , suggesting viruses appeared at a time when this early lineage was coalescing ( see below ) . Forcing molecules belonging to two superkingdoms into separate monophyletic groups once again confirmed the early split of Archaea and the late onset of Eukarya; the most parsimonious solution [constraint ( ( A ) ( E ) ) ] showed that the separate coalescence of the archaeal and eukaryal lineages followed the appearance of Eukarya as an organismal group [constraint ( E ) ] and always preceded the appearance of Bacteria [constraint ( B ) ] . Finally , constraining the three superkingdoms into separate monophyletic groups resulted as expected in higher S values , reflecting the coalescence of all lineages of a fully diversified organismal world . Out of all possible competing hypotheses of relationship several alternatives were most parsimonious , including an unresolved 3-taxon statement [constraint ( ( A ) ( B ) ( E ) ) ] . The confounding effects of recruitment were probably severe and were incapable of revealing the root of the universal tree at these high S values and late evolutionary stages . The timeline of organismal diversification provides evidence that the lineage of Archaea segregated from an ancient community of ancestral organisms and established the first organismal divide . The scenario of organismal diversification described above is congruent with our recent phylogenomic analyses of protein structure [16] and domain organization ( Wang and Caetano-Anollés , submitted ) in hundreds of completely sequenced genomes . The result is also congruent with recent studies that have used tRNA paralog ( alloacceptor ) clustering as a measure of ancestry of tRNA genotypes [40] and multiple lines of evidence [41] , [42] to suggest a Methanopyrus-proximal root of life . Although it is popularly accepted that the universal tree of life based on molecular phylogenies is rooted in the prokaryotes and that Archaea and Eukarya are sister groups , these recent results together with those presented in this paper offer compelling arguments in favor of an early appearance of the Archaea . Our evolutionary timeline is also remarkable in that it identifies three epochs in the evolution of the organismal world that were analogous to those proposed earlier [16]: ( 1 ) an architectural diversification epoch in which tRNA molecules diversified their structural repertoires ( light green areas in Figure 2 ) , ( 2 ) a superkingdom specification epoch in which tRNA molecules sorted in emerging lineages that specified superkingdoms Archaea , Bacteria , and Eukarya ( salmon areas ) , and ( 3 ) an organismal diversification epoch that started when all tRNA coalesced in each superkingdom ( light yellow areas ) . The evolutionary patterns observed in timelines appeared consistently in the absence or presence of class I or class II tRNA structural constraints ( Figure 2 ) . This suggests structural diversification predated organismal diversification during evolution of tRNA . Similarly , a scenario in which organismal diversification predated amino acid charging diversification was more parsimonious ( Table 3 ) , suggesting the discovery of both amino acid charging and associated codon function occurred in expanding lineages . These conclusions are supported by a recent study that compared sequence matches between tRNA halves and suggested the modern tRNA cloverleaf arose prior to the divergence of modern tRNA specificities and the three superkingdoms of life [32] . The organismal timeline inferred from tRNA sequence and structure showed Archaea was the most ancient superkingdom but established that viruses were also ancient . Viruses are relatively simple living entities and in many cases maintain a regular structure . They have long been considered fragments of cellular genomes and not living organisms and were generally excluded from consideration in evolutionary scenarios of the tripartite world , despite being important components of the biosphere . The importance of viruses and their potential roles in early cellular evolution were recently reevaluated [43] . A comparative analysis of structure and function , including virion assembly principles , suggested both RNA and DNA viruses may have been more ancient than previously thought , possibly even more ancient than the common ancestor of life [43] . However , they probably had a polyphyletic origin because structurally and functionally related viruses infect hosts in different lineages and even in different superkingdoms of the universal tree [44] , [45] . It is therefore possible that viruses form lineages and share a common ancestor , and that these lineages extend from the root to all branches in the tree of life . For example , the overall similarity of viral structures , such as coat protein folds enclosing nucleoprotein filaments , suggests a common mechanism for their appearance . The construction of phylogenies addressing the questions of origin and evolution of viruses in the context of the three superkingdoms are now possible with the increasing number of sequenced genomes of viral origin . In fact , comparative genomic analyses suggested viruses could be the source of new proteins for cells [46] . Many DNA informational proteins encoded today in cellular genomes probably originated in the viral world and were later transferred into the three cellular superkingdoms . Forterre recently proposed that DNA itself appeared in ancestral viral lineages [47] , [48] . He later on extended this proposal by suggesting that the DNA replication machineries of each superkingdom originated from three different ancestral viral lineages [49] . In his latest proposal , each cellular superkingdom originated independently from the fusion of an RNA-based cell and a large DNA virus [50] . In order to establish if the origin of the viruses was linked to one or more of the three superkingdoms of life we constrained viral and individual superkingdom tRNAs into competing monophyletic relationships ( Table 3 ) . Remarkably , most parsimonious constraints indicated viruses that associate with Eukarya and Bacteria had an origin in the archaeal lineage ( Figure 2 ) . The origin of viruses in Archaea is remarkable , especially if one considers the exceptional diversity and morphotype complexity of archaeal viruses [51] . Such an origin is compatible with the proposal by Forterre and colleagues that the transition from RNA to DNA genomes occurred in the viral world , and that cellular DNA and its replication machineries originated via transfers from DNA viruses to RNA cells . In fact , our phylogenomic analysis of structure [16] suggests a substantial portion of the replication machinery was developed during the architectural diversification phase immediately after reductive tendencies were already set in the archaeal lineage . This coincides with the relative time of emergence of viruses in the ancient world that was derived in this study . Since the appearance of a molecularly complex universal ancestor preceded the appearance of viruses , our results remain compatible with the accepted view that viruses originated from fragments of genetic material that escaped from the control of the cell and became parasitic ( the escape theory ) [52]–[55] . The origin of viruses is generally complex and may involve more than one mechanism [56] . Although several major classes of viruses are monophyletic , a common viral ancestry has not been evident [57] . Sequence analysis of viral genomes with various lengths ( ranging from a few to hundreds of kilobases and containing several to hundreds of genes ) and types ( ranging from double-stranded DNA to single-stranded RNA ) failed to reveal a common origin , suggesting instead polyphyletic ( multiple ) origins . However , a focus on sequence alone could be misleading . The viruses as a group contain more structural genomic diversity than cellular organisms such as plants , animals , or bacteria put together , and their sequences are fast evolving [58] . This could erase deep evolutionary history and confound analysis . Moreover , viruses also share many common features ( e . g . , genes coding for key proteins involved in viral replication and morphogenesis , parasitic nature of the replication mechanisms ) not shared by any kind of cellular organisms [57] , and these could be used to claim monophyly . This is especially true if the proposed ancient viral world existed [57] . This world harbored viral genes that retained their identity throughout the entire history of life . By this definition , the primordial pool of primitive genetic elements would be the ancestors of modern cellular and viral genes . This means that most , if not all , modern viruses were derived from elements that belonged to the primordial genetic pool , perhaps representing primitive form of self replicating DNA and precursor of life [59] . We end by noting that due to the small number of viral sequences sampled in our study , the conclusions drawn here should be taken with caution . However , a separate undergoing study analyzing a comprehensive dataset of tRNA sequences and structures but lacking information on base modifications support the evolutionary patterns presented in this study ( Ospina , Sun , and Caetano-Anollés , unpublished ) .
Part 2 ( compilation of tRNA sequences ) of the Bayreuth tRNA Database ( http://www . staff . uni-bayreuth . de/btc914/search/index . html; September 2004 edition; Table S1 ) contains a total of 571 tRNA sequences at RNA level with cloverleaf secondary structures . The structures were derived by comparative analysis using an alignment that is most compatible with tRNA phylogenies and known 3-dimensional models of structure [60] , [61] . The composition of part 2 was not pruned in our analyses and represents the most complete tRNA dataset currently available that contains information about base modifications . A total of 42 structural characters describing geometrical features of tRNA molecules ( Table S2 ) were scored , establishing character homology by the relative position of substructures in the cloverleaf [9] ( Sun and Caetano-Anollés , submitted ) . The length ( the total number of bases or base pairs ) and number of the substructures were coded as character states and were defined in alphanumerical format with numbers from 0 to 9 and letters from A to F . The minimum state ( 0 ) was given to missing substructures . We followed the Bayreuth database to treat the modified bases as deviations from the cloverleaf model . They were not allowed to establish canonical Watson-Crick pairs . Each helical stem region was scored as two complementary sequences ( 5′ and 3′ sides ) . The dataset was then partitioned into four subsets categorized by molecules belonging to each of the three superkingdoms or viruses/bacteriophages . In this study , a “total evidence” approach [62] , [63] ( also called “simultaneous analysis” [64] ) was invoked in phylogenetic analysis to combine both sequence and structure data of the complete ( 571 tRNAs ) and partitioned matrices . The goal of this analysis was to provide stronger support for the phylogenetic groupings recovered from analyses of structural data . We treated structural features in molecules as phylogenetic multi-state characters with character states transforming according to linearly ordered and reversible pathways . Character state transformations were polarized by assuming an evolutionary tendency towards molecular order . Characters were analyzed using maximum parsimony ( MP ) , a popular phylogenetic optimization method that searches for solutions that require the least amount of change . It is appropriate to treat geometrical features as linearly ordered characters because RNA structures change in discrete manner by addition or removal of nucleotide units . This causes gradual extension or contraction of geometrical features . Although insertion and deletion are also possible , they are more costly . The validity of character argumentation has been discussed in detail elsewhere [9] , [17] , [18] , [20] . A considerable body of evidence supports our polarization hypothesis depicting generalized trends applied to the structure of molecules: ( i ) the study of extant and randomized sequences shows that evolution enhances conformational order and diminishes conflicting molecular interactions over those intrinsically acquired by self-organization [20] , [65]–[70] , ( ii ) a molecular tendency towards order and stability has been experimentally verified using thermodynamic principles generalized to account for non-equilibrium conditions [71]; ( iii ) a large body of theoretical evidence supports the structural repertoire of evolving sequences from energetic and kinetic perspectives [72]–[74] , with some important predictions confirmed experimentally [75] , ( iv ) phylogenies generated using geometrical and statistical structural characters are congruent [9] , [20] , [21] , and ( v ) the reconstructions of rooted trees generated from sequence , structure , and genomic rearrangements at different taxonomical levels are congruent [17] , [18] , [20] , [21] , [76]–[78] . Phylogenetic trees were polarized by distinguishing ancestral states as those thermodynamically more stable . This results in reversible character transformation sequences that are directional and show asymmetry between gains and losses . Maximum and minimum character states were defined as the ancestral states for structures that stabilize ( stems , modified bases , and G:U base pairs ) and destabilize tRNAs ( bulges , hairpin loops , and other unpaired regions ) , respectively . All data matrices were analyzed using equally weighted MP as the optimality criterion in PAUP* v . 4 . 0 [79] . Because MP may outperform maximum likelihood ( ML ) approaches [34] , [35] , the use of MP is particularly appropriate for our analysis . ML is precisely MP when character changes occur with equal probability but rates vary freely between characters in each branch and when using large multi-step character state spaces ( decreasing the likelihood of revisiting a same character state on the underlying tree ) . This makes MP statistically consistent . Reconstructions of MP trees were sought using heuristic search strategies; 1 , 000 heuristic searches were initiated using random addition starting taxa , with tree bisection reconnection ( TBR ) branch swapping and the MulTrees option selected . One shortest tree was saved from each search . Hypothetical ancestors were included in the searches for the most parsimonious trees using the Ancstates command . BS values [80] were calculated from 105 replicate analyses using “fast” stepwise addition of taxa in PAUP* . The g1 statistic of skewed tree length distribution calculated from 104 random parsimony trees was used to assess the amount of nonrandom structure in the data [81] . Constraint analysis restricts the search of optimal trees to pre-specified tree topologies defining specific monophyletic groups , and was used here to test alternative or compare non-mutually exclusive hypotheses . The number of additional steps ( S ) required to force ( constrain ) particular taxa into a monophyletic group was examined using the “enforce topological constraint” option of PAUP* . The additional steps define an evolutionary distance that can be use to test alternative phylogenetic hypotheses or to compare hypotheses that are not mutually exclusive . The latter approach was used to construct evolutionary timelines , in which lower S values corresponded to ancient tRNAs , a trend that was derived from the rooted trees ( and embedded assumptions of polarization ) . Constraint analyses were conducted based on amino acid specificity or grouping of molecules by organismal superkingdoms or viruses . | The origins of the three major cellular lineages of life—Archaea , Bacteria , and Eukarya—and of viruses have been shrouded in mystery . In this study , we focus on transfer RNA , an ancient nucleic acid molecule that takes center stage in the process of protein biosynthesis and can be found everywhere in life . In a process that reconstructs history from molecular sequence and structure and at the same time forces molecules belonging to lineages into groups , we tested alternative hypotheses of origin and established when major organismal lineages appeared in evolution . Remarkably , timelines showed that Archaea was the most ancient lineage on earth and that viruses originated early in the archaeal lineage . Our findings unroot the universal tree of life , and , for the first time , provide evidence for an evolutionary origin of viruses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry/rna",
"structure",
"computational",
"biology/comparative",
"sequence",
"analysis",
"computational",
"biology/evolutionary",
"modeling",
"evolutionary",
"biology/bioinformatics",
"computational",
"biology",
"evolutionary",
"biology"
] | 2008 | Evolutionary Patterns in the Sequence and Structure of Transfer RNA: Early Origins of Archaea and Viruses |
The polymerization of actin in filaments generates forces that play a pivotal role in many cellular processes . We introduce a novel technique to determine the force-velocity relation when a few independent anchored filaments grow between magnetic colloidal particles . When a magnetic field is applied , the colloidal particles assemble into chains under controlled loading or spacing . As the filaments elongate , the beads separate , allowing the force-velocity curve to be precisely measured . In the widely accepted Brownian ratchet model , the transduced force is associated with the slowing down of the on-rate polymerization . Unexpectedly , in our experiments , filaments are shown to grow at the same rate as when they are free in solution . However , as they elongate , filaments are more confined in the interspace between beads . Higher repulsive forces result from this higher confinement , which is associated with a lower entropy . In this mechanism , the production of force is not controlled by the polymerization rate , but is a consequence of the restriction of filaments' orientational fluctuations at their attachment point .
Polymerization and depolymerization of microtubules or actin filaments , in the absence of any molecular motors , generate forces that are relevant to cellular processes , like cell membrane protrusion and propulsion of intracellular pathogens or organelles [1]–[8] . The energy is provided by the difference in chemical potential between the monomers ( G-actin ) in solution and the subunits incorporated in the filaments . The filament growth should slow and eventually stall as an opposite applied force approaches the thermodynamic limit given by [9] ( 1 ) where kBT is the thermal energy , δ , the elongation distance for the insertion of a new monomer ( 2 . 7 nm for actin ) , C , the concentration of monomers in solution , and Ccrit , the critical concentration for polymerization . A good estimate of this limit , at physiological concentration , would be a few piconewtons per filament . Actin filaments in close proximity to a load are thought to elongate through a ratcheting mechanism in which thermal fluctuations , either of the filament end or of the load , allow the stochastic insertion of new monomers in spite of a counteracting force [10] , [11] . The theoretical basis of actin-polymerization-generated forces is well developed for a single filament [12] or for a large ensemble of filaments described as a continuous material [13] . However , the individual behavior of filaments in an assembly has only been addressed in numerical simulations and is still under debate [14]–[16] , particularly how the macroscopic force is distributed on each filament . Experimental progress has been relatively slow . Convincing experiments have already been reported about the stalling force exerted by single growing actin filaments [17] as well as bundles of filaments [18] . At a much larger scale , the force-velocity profile generated by the growth of a densely branched network comprising thousands of filaments has been measured by several groups [19]–[22] . These experiments give insights into what happens in cells but are too complex to give information on a microscopic mechanism . The force-velocity profile of a controlled , small number of actin filaments has not been measured yet because such experiments require handling short filaments and controlling their organization . The force-velocity profile should be very informative , since its shape is dictated by the microscopic mechanism through which the chemical energy is transduced into force . In this paper we present the force-velocity profile measured for a few actin filaments ( typically 10 to 100 filaments ) using an original setup . We demonstrate that the force , in our geometry , is due to the entropic restriction of the rotation of the growing filaments at the anchoring points .
Here , we have designed an experiment that allows simultaneous measurements of the growth velocity , the loading force , and the elastic response of a few growing filaments . We use 1 . 1-µm-diameter magnetic colloids that form linear chains when a magnetic field is applied . The distance between the colloids and the magnetic attractive force is accurately monitored through the application of a controlled homogeneous magnetic field . Typically , the force can vary from 0 . 1 pN to almost 100 pN , while the distance X between colloidal surfaces can vary from a few nanometers to several micrometers . The magnetic beads are functionalized with gelsolin , a strong capping protein of actin filament barbed end , at a controlled surface density . The total number of active gelsolins per bead , NGS , is measured separately . In our experiments we use NGS = 4 , 000 or 10 , 000 , corresponding to mean distances of 33 and 21 nm , respectively , between the gelsolins anchored on the colloids . A detailed characterization of the colloid surface chemistry and a detailed description of our experimental setup are given in Materials and Methods . In the presence of G-actin , gelsolin initiates pointed end growth of filaments at the surface of the beads , as shown schematically in Figure 1A . The radial growth of gelsolin-anchored filaments away from the bead surface causes an increase in the bead-to-bead distance , visualized in video-microscopy images ( Figure 1B and Video S1 ) . This qualitative observation clearly indicates that the actin polymerization induces forces larger than a few piconewtons . More precisely , G-actin is first added to the suspension of gelsolin-functionalized magnetic beads at time t = 0 . Before the application of the magnetic field at t = t0 , the filaments grow freely with a pointed end growth rate v0 up to a length L0 . Figure 1C shows the plot of the beads' center-to-center distance d as a function of subsequent time , with constant loading forces ranging from 0 . 5 pN to 35 pN . d increases linearly with time for each force , and therefore the beads' separation velocity vbead can be precisely measured as a function of applied magnetic force , f . At low forces , this velocity tends to the actin growth velocity in solution derived from kinetics measurements ( v0 = 0 . 42 nm/s; see Materials and Methods ) . When the force is increased , the velocity decreases , reaching almost zero at 35 pN ( Figure 1D ) . In a second experiment , we explored the link between the beads' velocity and the filaments' elongation rate . To address this question , a sequence of low-high-low forces is applied to a chain ( Figure 2 ) . During high-force application , the distance d between beads is almost constant . However , when the force is suddenly released , the distance increases very rapidly to a value that is higher than the one at the end of the first low-force period ( Δd = 63 nm ) . This result shows that filaments must have grown during the high-force application . Indeed , the observed response is too rapid to be attributed to actin repolymerization . Moreover , the distance versus time before and after application of high force collapse on the same line . This result strongly suggests that the filaments grow as though they were in solution whatever the value of the applied force . In the experiments , the velocity at which the beads separate under force is not directly the velocity at which the actin filaments grow . It appears from the previous results that , in our experiments , filaments are not just pushing the beads perpendicular to their surface; their organization in the contact zone is much more complex . Filaments are too short to be directly observed with an optical microscope , so we investigated the detailed mechanical response of our system in an indirect fashion . The instantaneous force-distance profile between colloidal particles within a chain [23] , [24] can indeed be obtained using the present setup . At mechanical equilibrium , the net force applied to each bead within the chain is zero , so the attractive magnetic force is strictly balanced by a repulsive force . Hence the repulsive force-distance profile is reconstructed by measuring the interparticle distance X between the beads' surfaces at different applied magnetic forces f , from 0 . 1 to 50 pN . We carry out the experiment fast enough to neglect the filament growth . In Figure 3 we present the applied force f as a function of X . This force-distance profile is measured for two different values of L0 , 200 and 400 nm , over a cycle of compression and decompression . There were three important results of this experiment . First , the mechanical response depends on the initial filaments' lengths . Second , for each length , the two branches of the cycle appear to be equivalent: the force profile is reversible ( at low forces the thermal energy induces large fluctuations of the interparticle distance ) . This indicates that the filaments that build up the repulsive force respond elastically to the load without being damaged by the compressive jump . Finally , as already observed in Figure 2 during switches of the force , the deformation is very high ( [X − Xf = 0]/Xf = 0 >50% ) : the elastic modulus of the contact assembly is very small ( ∼1–10 Pa ) . We will now demonstrate that this soft elasticity is a consequence of the orientational fluctuations of the filaments . The first step is the evaluation of the number of filaments that build up the elastic response . We first evaluated the total number of filaments per bead , NGS ( Materials and Methods ) . The number N of filaments that can sense the opposite surface is estimated based on a geometrical argument that is illustrated in Figure 4A: a filament of length L is counted only if it can touch the opposite bead . We have assumed that the filaments are inter-digitated . This hypothesis is supported by the low concentration of anchoring points and by the fact that the bead velocity at zero force equals the filament's growth velocity in solution ( Figure 1 ) : in a tip-to-tip geometry , the distance between beads should increase twice as fast as the filament length . For filaments of length L , for a surface-to-surface separation X , and for a total number of filaments per particle NGS , the number , N , reads ( 2 ) where R is the radius of particles . N typically ranges from 0 to 250 when X is changed from 400 nm down to a few nanometers , for NGS = 4 , 000 and L = 400 nm . We now consider the mechanism that drives the observed soft mechanical behavior ( Figure 3 ) . Buckling of filaments is ruled out by their small length ( L<500 nm ) . Indeed , a lower estimation of the buckling force Fbuck is the critical Euler force for one filament [12] multiplied by the number of filaments that should buckle ( 3 ) Lp being the persistence length ( about 8 µm for actin [25] ) . Fbuck is about two orders of magnitude larger than the measured force . Since filaments can be truly considered as short rigid rods , we alternatively consider the possibility that elasticity of the system arises from the link between biotin and gelsolin that anchors the filaments to the beads: biotin is bound to gelsolin via a reactive group that contains a flexible spacer of 1 . 35 nm ( see Materials and Methods ) . This link may act as a free molecular hinge that allows the filaments to be tilted without bending as the opposite surface approaches , as schematized in Figure 4B . Moreover , the free hinge gives rotational degrees of freedom to the filaments . The elastic response is then hypothesized to be solely due to the entropic restriction imposed by the approaching surface . Indeed , the number of accessible configurations Ω is proportional to the surface area described by the filament tip , and Ω decreases with the confinement: if X>L , Ω ∝ 2πL2 , whereas Ω ∝ 2πLX if X<L ( Figure 4B ) . From the free energy F = −kBT ln Ω , one can compute the repulsive force . This force is null when there is no confinement , i . e . , X >L , and is given by fev = kBT/X when X<L . We then simply assume that in our case the total force due to N filaments will be given by c N ( X , L ) fev , where c is a coefficient that accounts for geometrical effects , i . e . , the non-planar surface of the beads . Using the value of N given by Equation 2 , the forces calculated for the two experimental values of L0 are compared to the measured ones as shown in Figure 3 . The scaling matches the data well , with c = 0 . 2±0 . 1 as the sole adjustable parameter for both curves . We show here that the force-velocity profile is also governed by the entropic repulsion . If the bead-to-bead distance is kept constant , more filaments are confined as they elongate , and thus the repulsive force increases . In our experiments the force ( and not the bead-to-bead distance ) is kept constant during filament growth; hence , when the filaments elongate , the distance is modified in a way that keeps the force constant . If the force is exclusively entropy-driven , since L ( t ) = v0 t , the bead-to-bead distance X can be calculated from an expansion of N ( X , L ) at first order in X/R and L/R , as follows: ( 4 ) The assumption that force is entropy-driven yields a linear increase with time of the bead-to-bead distance , in agreement with experimental data ( Figures 1C and 2 ) . From Equation 4 , we get directly the velocity: ( 5 ) To further test our model , we performed experiments at different forces and different total numbers of filaments ( Figure 5 ) . We also show the comparison of the force-velocity profiles measured and computed from Equation 5 for two values of NGS , 4 , 000 and 10 , 000: both the shape and magnitude match the data very well . The curve fitting gives c = 0 . 18±0 . 02 for NGS = 10 , 000 , and c = 0 . 13±0 . 02 for NGS = 4 , 000 . These c values are in good agreement with the c value obtained from the mechanical measurements ( Figure 3 ) . In addition , this model gives the right limit v0 = 0 . 42 nm/s for the velocity at f = 0 . Finally , according to this model , the velocity never becomes negative even at high forces . Consistent with this , the data actually show that no depolymerization was induced by the applied force .
The force transduction mechanism demonstrated here may be considered as an alternative to the Brownian ratchet model , in a case where large fluctuations in orientation of the growing filament occur as the elongating tip gets close to a surface . In the classical Brownian ratchet model , the growth of the filaments is slowed down by the load that is assumed to be applied to their tip [10] , but the organization of the filaments and their perpendicular orientation relative to the surface remain unchanged . In our experiment , the filament growth is not modified by the opposing surface , but filaments orient themselves on average by decreasing their angle to the surface of the magnetic bead with increasing loads . Here the repulsive force pushing the beads away from each other is due to the restriction of rotational freedom around the flexible streptavidin-biotinylated gelsolin linker , which decreases the entropy of the hinged filaments , while they still elongate as in solution . This is analogous to the osmotic pressure , but with orientational degrees of freedom and not translational ones . In addition , this entropy-driven mechanism develops significant forces , typically a few tenths of the theoretical maximum for one filament ( see Equation 1 ) . Does the present mechanism for force production by actin assembly have a physiological relevance in cell motility ? Actin arrays that support cell migration are generally oriented with their barbed ends abutting the leading edge , where new filaments are created either by nucleation and processive elongation by formins , or by branching via the WASP/Arp2/3 machinery . Daughter filaments initiated by Arp2/3 branching grow at a 70° angle from the mother filament . Although the branch junction allows some flexibility in orientation of the daughter filament when branched filaments are formed in solution [26] , it likely behaves as a more rigid hinge than the streptavidin-gelsolin link in our experiments . The rigidity must also be enhanced by the constraints imposed in the context of the intricate dendritic lamellipodial array [27] . However , some cases exist in which the concepts presented here for force production may apply . Migrating cells actually display a variety of phenotypic morphologies of the lamellipodium [28] . In rapid cell migration of keratocytes , the turnover of a densely branched array feeds fast protrusion associated with a persistent smooth morphology of the leading edge . Filaments keep a constant orientation toward the front , in part because of the interaction of barbed ends with membrane-associated regulators like VASP , which maintain some processive link with growing barbed ends [28] , [29] . In the absence of such links , cells migrate with lower directional persistence and the leading edge adopts variable shapes . Within our model , this phenotype may be generated by the greater freedom of reorientation experienced by filaments that would be present in lower number . Similarly , the protrusive activity appears to vary in correlation with variable angles of the filaments to the cell front in the range 15° to 90° [30] . Interestingly , Koestler et al . [30] observed a correlation between the protrusion rate and the filament orientation , similar to our in vitro observations . Hence , depending on barbed end regulation , filament density , and velocity of protrusion , different mechanisms of force production by actin assembly may be at work in migrating cells , and some room may be found for a physiological role of the change in filament orientation in force production . In conclusion , if the key components at play for cell motility are clearly identified , how their temporal and spatial organizations are regulated in motile processes is still to be unraveled . We believe that our novel experimental approach provides clues to achieve this goal .
Actin was purified from rabbit muscle as previously described [31] and isolated as Ca-ATP-G-actin by Superdex-200 chromatography [32] in G-buffer ( 5 mM Tris [pH 7 . 8] , 0 . 1 mM CaCl2 , 0 . 2 mM ATP , 1 mM DTT , 0 . 01 wt% NaN3 ) . Ca-actin was converted to Mg-actin by incubation in 0 . 2 mM MgCl2 and 0 . 25 mM EGTA just before experiments . Actin was pyrenyl labeled as previously described [33] . Recombinant human gelsolin was expressed and purified as previously described [34] , and stored in Tris buffer at −80°C ( 20 mM Tris [pH 7 . 5] , 1 mM EGTA , 0 . 15 M NaCl2 , 0 . 01 wt% NaN3 ) . Protein was first dialyzed in a PBS buffer with 1 mM EGTA and 0 . 01 wt% NaN3 ( pH 7 . 5 ) , then biotinylated with sulfo-NHS-biotin ( EZ-Link Sulfo-NHS-Biotin Reagents , spacer arm length 1 . 35 nm; Thermo Scientific ) during 45 min at room temperature . Biotinylated gelsolin was used immediately after preparation . To determine the molar ratio of biotin to protein , the HABA method was used ( Pierce Biotin Quantitation Kit; Thermo Scientific ) and gave 15 biotins per gelsolin . The critical concentration Ccrit and the growth rate kon at the pointed end in our salt conditions were both derived from pyrene fluorescence assays [35] . Fluorescence polymerization measurements were performed using 2 µM monomeric actin ( 10% pyrenyl-labeled ) in the polymerization buffer ( 5 mM Tris , 40 mM KCl , 0 . 6 mM MgCl2 , 0 . 2 mM CaCl2 , 0 . 2 mM ATP , 1 mM DTT , 0 . 5 wt% F-127 , 0 . 01 wt% NaN3 [pH 7 . 8] ) . Ccrit = 0 . 7 µM was given by the equilibrium concentration of serially diluted F-actin samples in the presence of gelsolin . kon was assayed from kinetic measurements using 1:2 gelsolin-actin ( GA2 ) complexes at different concentrations . Fitted exponential curves give kon = 0 . 12 µM−1·s−1 . From our working concentration of monomeric actin ( C = 2 µM ) , we computed the polymerization velocity from the pointed end in solution , v0 = konδ ( C − Ccrit ) = 0 . 42 nm/s , and stalling force for a single filament , Fstall = ( kBT/δ ) ln ( C/Ccrit ) = 1 . 6 pN . Streptavidin-coated superparamagnetic particles of 1 . 135 µm in diameter ( 5 µl ) ( Dynabeads MyOne Streptavidin , 106 streptavidins per particle; Dynal-Invitrogen; size measured by dynamic light scattering method ) were washed five times in Tris buffer containing the pluronic surfactant F-127 ( 5 mM Tris [pH 7 . 5] , 0 . 5 wt% F-127 , 0 . 01 wt% NaN3 ) . Then , they were incubated with 1 µM freshly biotinylated gelsolin for 10 min for the maximal density . Next , 0 . 5 mM biotin is added to the sample to block streptavidin free sites . After 5 min , the grafted particles were washed five times with 5 mM Tris ( pH 7 . 8 ) , 0 . 2 mM CaCl2 , 0 . 5 mM biotin , 0 . 2 mM ATP , 1 mM DTT , 0 . 5 wt% F-127 , 0 . 01 wt% NaN3 , and stored in the same buffer . A fraction of the grafted particles were used to determine the grafting density ( see below ) . Next , 0 . 01 wt% grafted particles were mixed with 2 µM monomeric actin in the polymerization buffer . The obtained solution was rapidly transferred into a capillary tube ( Vitrocom ) that was sealed at both ends and attached to a slide with liquid wax . Force-velocity measurements and gelsolin biotinylation were performed on the same day . One sample was used for one point in the velocity-force diagram , and up to ten samples could be made with one gelsolin preparation . The total number of active gelsolins per bead , NGS , was obtained from pyrene-actin fluorescence assay . Actin polymerization was induced by adding 0 . 01 wt% gelsolin-coated beads to 2µM G-actin in the polymerization buffer . The number of growing filaments , and therefore NGS , was computed from kinetic measurements using the measured association rate at the pointed end: kon = 0 . 12 µM−1·s−1 . We used NGS of 4 , 000 or 10 , 000 in our experiments , corresponding to a mean distance between gelsolins of 33 nm and 21 nm , respectively . Before force measurements , filaments were allowed to grow freely on the beads until t0 . The initial length of the filaments L0 was directly deduced from this time , L0 = v0 t0 , where v0 is the pointed end growth velocity in solution . The chains of typically ten beads were imaged using a Nikon TE-2000 inverted optical microscope . Two electromagnetic coils , mounted onto a motorized stage ( ECO-STEP; Märzhäuser ) , generated a magnetic field from 0 to 100 mT . Images were collected through a 100× oil immersion objective ( N . A . 1 . 25 ) using a digital camera ( ORCA-ER; Hamamatsu ) . The bead positions were obtained with a particle tracking algorithm using NIS-Elements Nikon software [36] . The mean interparticle distance was then calculated by averaging the distance between particles within the chain , and the magnetic force was calculated from the magnetic field and the mean distance d according to [24] . Electrostatic repulsive forces between beads under our conditions ( 40 mM KCl ) are negligible as compared to our measured forces . For instantaneous force-distance profiles , a cycle of compression-decompression takes about 2 min to complete , allowing about 50 nm of growth for the filaments . For velocity-force profiles , the mean distance was measured every 15 s , and the magnetic field was adjusted in order to keep the magnetic force constant . The experiments were performed with several purifications of actin , two different batches of magnetic beads , and two different purifications of gelsolin , without noticeable difference . Velocity-force curves are shifted to lower velocity if the actin preparation is too old ( more than 3 wk ) . For force-velocity measurements ( Figure 5 ) , the main variability is coming from the grafting procedure , inducing changes in the gelsolin density . Measurements are more reproducible within the same batch of grafted beads ( six points from one preparation at 7 . 3±0 . 1 pN , vbead = 0 . 076±0 . 011 nm/s; ten points from two preparations , f = 7 . 2±0 . 5 pN , vbead = 0 . 126±0 . 04 nm/s ) , but the whole profile cannot be obtained with a single preparation . For force-distance measurements ( Figure 3 ) , the main source of error is the polydispersity of the beads ( 1% in size ) , since we are using typically ten beads in a chain . From one chain to another , the force profiles are identical , but can be shifted on the x-axis by typically 10 nm . We indeed removed the mean bead diameter to obtain X from the center-to-center measurements . This 10-nm shift in X impacts the c value obtained by fitting the data; an error of 0 . 1 for c is a conservative estimation . | Actin self-assembles into filaments , and this produces forces that deform cell membranes in a large number of motile processes . While physical measurements have been performed of the force produced by growth of either a single filament or a large intricate array of filaments organized in an active macroscopic gel , these measurements don't provide a clear picture of how force is produced by the assembly of each filament within a complex structure . The present study explores a situation between these two extremes by measuring the force produced by the assembly of a small number of filaments . We developed a method in which actin filaments grow from the surface of magnetic beads that are aligned by a controlled magnetic field . The distance between beads in a chain-like arrangement increases with time when the force is kept constant . We observe that the growth of actin filaments is not affected by the load , in contrast to the widely accepted “Brownian ratchet model . ” Instead , our results suggest that the surface opposite growing filaments imposes restrictions on the rotational fluctuations of a filament at its free hinge anchoring point , inducing a repulsive force . The confinement of filaments increases as they grow , and this in turn increases the repulsive force developed by their growth . This entropy-based mechanism may operate during motile processes when actin networks are loosely organized . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biophysics/macromolecular",
"assemblies",
"and",
"machines",
"biochemistry/macromolecular",
"assemblies",
"and",
"machines",
"biochemistry/experimental",
"biophysical",
"methods",
"cell",
"biology/cytoskeleton"
] | 2011 | Force-Velocity Measurements of a Few Growing Actin Filaments |
Plants continuously extend their root and shoot systems through the action of meristems at their growing tips . By regulating which meristems are active , plants adjust their body plans to suit local environmental conditions . The transport network of the phytohormone auxin has been proposed to mediate this systemic growth coordination , due to its self-organising , environmentally sensitive properties . In particular , a positive feedback mechanism termed auxin transport canalization , which establishes auxin flow from active shoot meristems ( auxin sources ) to the roots ( auxin sinks ) , has been proposed to mediate competition between shoot meristems and to balance shoot and root growth . Here we provide strong support for this hypothesis by demonstrating that a second hormone , strigolactone , regulates growth redistribution in the shoot by rapidly modulating auxin transport . A computational model in which strigolactone action is represented as an increase in the rate of removal of the auxin export protein , PIN1 , from the plasma membrane can reproduce both the auxin transport and shoot branching phenotypes observed in various mutant combinations and strigolactone treatments , including the counterintuitive ability of strigolactones either to promote or inhibit shoot branching , depending on the auxin transport status of the plant . Consistent with this predicted mode of action , strigolactone signalling was found to trigger PIN1 depletion from the plasma membrane of xylem parenchyma cells in the stem . This effect could be detected within 10 minutes of strigolactone treatment and was independent of protein synthesis but dependent on clathrin-mediated membrane trafficking . Together these results support the hypothesis that growth across the plant shoot system is balanced by competition between shoot apices for a common auxin transport path to the root and that strigolactones regulate shoot branching by modulating this competition .
Plants can alter their body plan to adapt to the environment in which they are growing ( reviewed in Leyser 2009 [1] ) . This is possible because plant development is continuous , with postembryonic development being dependent on the activity of meristems . For example , the primary shoot apical meristem is laid down during embryogenesis at the apical embryonic pole , and after germination , the meristem gives rise to the adult shoot system through the production of a series of phytomers consisting of a leaf , a segment of stem , and a new shoot apical meristem , established in the axil of each leaf . These axillary meristems can remain dormant , or they can activate to produce a new shoot axis , with the same developmental potential as the primary shoot . Thus the mature shoot system can range from a solitary stem to a highly ramified bush , depending on the activity of the axillary meristems . The large number of meristems in the shoot system allows the plant to recover quickly from damage and to adjust its growth according to spatially heterogeneous environmental inputs such as unilateral shading , and to systemic inputs such as the nutrient status of the plant . Thus multiple inputs are integrated to balance growth across the shoot system . Axillary meristem activity is controlled by a network of systemically moving endogenous signals , among which auxin plays a pivotal role . Auxin , synthesized principally in the young expanding leaves of growing shoot tips , moves rootward in the stem through the polar auxin transport stream ( PATS ) . The direction of the PATS is determined by the polar localisation of PIN-FORMED ( PIN ) plasma membrane ( PM ) auxin efflux carriers [2] , [3] . In the stem , efficient rootward auxin flow requires PIN1 [4] , which is basally localised in the PM of xylem parenchyma cells [5] . Auxin in the PATS inhibits the outgrowth of axillary buds . Pharmacological inhibition of the PATS or removal of the primary shoot apex triggers outgrowth of axillary buds , and application of auxin to the decapitated stump prevents this outgrowth [6] , [7] . However , direct application of auxin to axillary buds does not prevent their outgrowth [8] , and apically applied auxin is not transported into the axillary buds [9] , suggesting that auxin in the PATS inhibits shoot branching indirectly . Two nonexclusive mechanisms have been proposed to account for the indirect action of auxin . Firstly , it has been proposed that auxin regulates the synthesis of one or more second messengers , which move up into the axillary buds to regulate their activity . Two classes of phytohormone , cytokinins and strigolactones , are good candidates for these signals . Cytokinins can move up the plant in the transpiration steam in the xylem . Direct application of cytokinin to axillary buds can induce outgrowth , even in an intact plant [10] . Decapitation elevates but auxin application reduces endogenous cytokinin levels in xylem sap [11] and in the stem of nodal explants [12] . Together these data suggest that auxin inhibits bud outgrowth in part by reducing systemic and local cytokinin levels , and thus cytokinin supply to buds . A similar dataset exists for strigolactones . They can also be transported up the plant in the xylem [13] . Their direct application to buds can inhibit outgrowth on intact and decapitated plants [14] , and decapitation reduces but auxin application elevates the transcription of strigolactone biosynthetic genes [15] , [16] . These data suggest that auxin inhibits bud activity in part by increasing systemic and local strigolactone synthesis and thus strigolactone levels in buds . However , strigolactones only inhibit shoot branching in the presence of a competing auxin source , such that supply to a solitary bud has little or no effect and supply to an explant carrying two buds inhibits only one of the buds , which can be either the more apical or more basal bud [17] , [18] . Furthermore , strigolactone addition results in a reduction in PIN1 levels in xylem parenchyma cells within 6 h , accompanied by a reduction in polar auxin transport [17] . Thus in strigolactone biosynthetic mutants , high levels of branching correlate with high levels of PIN1 and polar auxin transport and high auxin concentration in the main stem [19] , [20] . These observations have led to an alternative model both for strigolactone action and for the indirect mode of inhibition of axillary bud growth by auxin in the PATS in the main stem . This alternative model derives from considerations of the auxin transport canalization hypothesis , originally proposed to explain vascular pattern formation . The central tenet of the canalization hypothesis is positive feedback between auxin flux and auxin flux capacity [21] . Restated in terms of PIN proteins , an initial passive flux of auxin from an auxin source to an auxin sink results in the up-regulation and polarisation of PINs in the direction of the flux . This results in files of cells with high levels of PINs polarised in the direction of the sink , some of which may differentiate into vascular strands . The emergence of such files between an auxin source and sink has been directly observed [22] , [23] . Given that active axillary buds are sources of auxin [23] , [24] , and the main stem can act as an auxin sink , by carrying auxin away to the root , auxin transport canalization can act to connect the bud to the stem , transporting auxin away from the bud apex and establishing vascular connectivity between the bud and the rest of the plant . However , high auxin levels in the main stem can prevent canalization of auxin transport out of the bud by reducing stem sink strength for auxin , limiting the initial flux of auxin out of the bud , thereby preventing escalation of the positive feedback at the heart of the canalization process [20] , [23] . If auxin transport canalization out of the bud is required for bud activity , then this could explain the indirect inhibition of buds by auxin in the main stem , without the need for a second messenger relaying the auxin signal into the bud . Instead , buds and the main shoot apex compete for access to a common auxin transport pathway down to the root . Computational simulations of this model demonstrate its plausibility [20] . Moreover the model can explain the association of high branching with high PIN1 , auxin transport , and auxin levels observed in strigolactone mutants , by postulating that the mode of action of strigolactone is to reduce the accumulation of PIN1 on the PM , thus making canalization more difficult to achieve , requiring a higher initial flux of auxin to drive the positive feedback loop . The model also explains the requirement for a competing auxin source for strigolactone-mediated bud inhibition [17] , [18] . One attractive feature of this model is that it establishes a regulatory framework underpinning the ability of plants to balance growth across the shoot system , integrating local ( e . g . , light quality ) and systemic ( e . g . , nutrient limitation ) information , through bud–bud competition . However , the validity of this model remains controversial because of the substantial body of evidence consistent with the hypothesis that strigolactones act locally and specifically in buds to inhibit their growth , by up-regulating the transcription of the TCP family transcription factor , BRC1 , which is known to be required for stable bud inhibition [14] , [25] , [26] . In this article , we use computational modelling to generate predictions that allow these alternative hypotheses for strigolactone action to be distinguished . Our results strongly support the auxin transport canalization model for shoot branching control . Specifically , we demonstrate that strigolactone treatment can either inhibit or promote shoot branching , depending on the auxin transport status of the treated plants . This is difficult to reconcile with direct bud inhibition by strigolactone . In contrast , these responses can be explained if strigolactones act by regulating PIN1 removal from the PM of cells in the shoot . Consistent with this mode of action , we show that a rapid primary response to strigolactone is clathrin-dependent PIN1 depletion from the PM .
Many biological behaviours are the outcome of interlinked feedback regulation acting recursively . Consequently , these behaviours are difficult to understand by intuitive interpretation of biological observations . Formalisation of these systems through mathematical modelling and computer simulation can link mechanistic hypotheses for their action to emergent higher order behaviour and thereby increase understanding of the underlying mechanisms . We previously presented a computational model for shoot branching control , based on the auxin transport canalization hypothesis described above [20] . This model can account for the phenotypes of gn or tir3 mutation , and strigolactone treatment , if their actions are to reduce insertion or enhance removal of PIN1 from the PM [20] . The heart of the model is Equation 1 , which encapsulates the positive feedback of auxin transport canalization . PIN1 levels in the membrane depend on both insertion , captured by a rate ( ρ ) proportional to the flux of auxin across the membrane , and removal , captured by a rate ( μ ) ( for full details , see [20] ) : ( 1 ) To dissect further which parameters in this model might be affected by max , gn , and tir3 mutation , we set “wild-type” values of the parameters and ran simulations with individual input values for each parameter in turn , changed around the wild-type value . The simulation outputs are summarised for shoot branching levels , polar auxin transport levels , and PIN protein levels in Table 1 . Of the 14 parameters , 13 were able to capture branchy phenotypes with some input values . Of these , only three captured both branchy phenotypes and altered levels of polar auxin transport . These were ρ ( the PIN insertion constant ) , μ ( the PIN removal constant ) , and T ( the polar transport coefficient—the efficiency with which each PIN protein transports auxin ) . To match the biological data , GN and TIR3 activity should be explained by a parameter whose reduction can elevate branch numbers , reduce polar auxin transport , and reduce PIN1 accumulation ( Figure 1 ) . Only ρ ( the PIN insertion constant ) satisfies these criteria ( Table 1 ) . Similarly , strigolactone/MAX activity should be explained by a parameter whose reduction can increase shoot branching , polar auxin transport , and PIN1 accumulation ( Figure 1 ) . Only μ ( the PIN removal constant ) satisfies these criteria ( Table 1 ) . To understand better the relationship between the parameters and simulation outputs , we plotted two 3-dimensional graphs that show PAT ( Figure 2A ) and shoot branching ( Figure 2B ) levels as heights on the μ–ρ plane . The relationship between polar auxin transport levels and μ–ρ was relatively simple: as PIN removal ( μ ) decreased and PIN insertion ( ρ ) increased , the polar auxin transport level gradually increased , resulting in a smooth slope ( Figure 2A , C , D ) . In contrast , the relationship between shoot branching level and μ–ρ was more complex: as PIN removal ( μ ) decreased , the shoot branching level increased , creating a plateau of high branching at low μ values . However , as PIN insertion ( ρ ) decreased the branching level increased , even when PIN removal ( μ ) was quite high , resulting in a ridge of high branching ( Figure 2B ) . High branching on the low μ ( low PIN removal ) plateau is caused by easy establishment of canalization of auxin transport from bud to stem , with low initial auxin fluxes able to establish canalization through positive feedback , making buds difficult to inhibit . High branching along the low ρ ( low PIN insertion ) ridge is caused by low auxin efflux from active shoot apices , such that a larger number of active apices are needed to supply sufficient auxin to the main stem to prevent activation of further buds . The profiles for branch number at any one μ or ρ value made much more abrupt transitions than for auxin transport levels ( Figure 2C , D ) , with mostly high or low branch numbers , and only narrow regions of parameter space giving intermediate branch numbers . This is because branch activation in the model is triggered by canalization of auxin transport out of the simulated bud and the positive feedback inherent in the canalization process produces switch-like behaviour [20] . To capture the behaviour of strigolactone biosynthesis mutants such as max4 or strigolactone-signalling mutants such as max2 , we assigned a low value to μ , conditioning slow PIN removal . This resulted in higher levels of both polar auxin transport and branching compared with those of the defined wild-type ( Figure 2A , B ) , consistent with biological results ( Figure 1 and [17] , [19] ) . Similarly we simulated the gn or tir3 mutations as a low ρ value , conditioning low PIN insertion , resulting in a lower level of polar auxin transport and a higher level of branching ( Figure 2A , B ) , as observed in biological experiments ( Figure 1 and [20] , [33] , [35] ) . To simulate addition of the synthetic strigolactone , GR24 , we increased the value of μ ( increasing PIN removal ) , which gave slightly lower polar auxin transport and shoot branching levels compared to the defined wild-type ( Figure 2A , B ) , consistent with published biological data [17] . When the low μ value of max and the low ρ value of gn or tir3 were simultaneously applied , the model predicts moderate polar auxin transport levels and high branching , consistent with biological results ( Figure 1 and [20] ) . Thus , single parameter changes in the model capture the phenotypes of wild-type , single and double mutants , and where known , their responses to GR24 . Furthermore , the relative magnitude of the responses to GR24 in different genetic backgrounds and with respect to branching versus auxin transport is also captured . This analysis led to an interesting and counterintuitive prediction . The dose-response curve of max4 tir3 branch number to GR24 is predicted to have two peaks , which lie on the low PIN removal ( μ ) plateau and low PIN insertion ( ρ ) ridge ( Figure 2B ) . To test this prediction , we grew wild-type , max4 , tir3 , and max4 tir3 plants for 8 wk on agar-solidified medium supplemented with GR24 ranging from 10 nM to 1 µM ( Figure 3A ) . As previously shown [17] , in both wild-type and max4 , GR24 reduced branching levels monotonically , although this effect was not statistically significant in the wild-type . In contrast , in tir3 , GR24 significantly elevated branching levels at 10 nM , and reduced branching at higher concentrations , with 1 µM resulting in very poor growth . In max4 tir3 , 10 nM GR24 reduced branching levels , but 50 nM GR24 restored branching levels to those of untreated plants and higher concentrations reduced them again . This latter part of the curve was shifted compared to the tir3 alone , with branched plants produced at 1 µM , a concentration that severely inhibits growth in tir3 mutants . Therefore , GR24 did not simply inhibit but also promoted shoot branching depending on the concentration and the genetic background of the treated plant . These results validate the predictions of the model with the minor modification that the effects of tir3 mutation on PIN insertion ( ρ ) suggest that it is placed on the low μ slope of the low ρ ridge , rather than at its summit , as proposed in Figure 2 . As well as the unusual dose–response relationships , the parameter space exploration predicts no branching at high PIN removal ( μ ) and low PIN insertion ( ρ ) , caused by insufficient auxin transport to support bud growth . In the dose–response experiments described above , 1 µM GR24 severely affected the growth of the tir3 mutant . To explore the response of tir3 and gn mutants to high levels of GR24 in more detail , we grew wild-type , max4 , gn , and tir3 plants for 8 wk on agar-solidified medium containing 5 µM GR24 , or an equivalent volume of solvent . GR24 affected the overall vigor of gn and tir3 plants , such that their total dry weights were significantly reduced compared to untreated controls ( Figure 3B , C ) . This effect was particularly noticeable in tir3 plants ( Figure 3B , C ) , which often did not survive to maturity in the presence of 5 µM GR24 . GR24 had no effect on dry weight in wild type or max4 ( Figure 3C ) . Thus gn and tir3 shoots are hypersensitive to GR24 . These data strongly support the hypothesis that strigolactones increase the removal of PIN1 from the PM , and indeed we have previously shown that GR24 treatment reduces PIN1 abundance in xylem parenchyma cells within 6 h in a MAX2-dependent manner [17] . To investigate the dynamics of this process in more detail , we prepared hand sections of stems of different genotypes harbouring the PIN1–GFP fusion , as described above , and recorded basal PM PIN1 levels every 10 min over a 90-min period . PIN1 was significantly reduced by the addition of 5 µM GR24 within 40 min in wild-type plants and within 30 min for max1 plants ( Figure 4A ) . As expected , no significant difference was observed in max2 mutants ( Figure 4A ) . We also examined wild-type sections treated with 50 µM cycloheximide for 30 min before a 60-min incubation with 5 µM GR24 or the vehicle control . GR24-induced depletion of PM PIN1 level was unaffected by cycloheximide treatment ( Figure 4B ) , suggesting that this process is independent of new protein synthesis . Depleted PM PIN1 could in principle result from either increased removal or reduced insertion of PIN1 . In roots , there is good evidence that many membrane proteins , including PIN1 , cycle rapidly between the PM and endomembrane compartments [28] . The removal of these proteins from the PM is mediated by clathrin-dependent endocytosis , which is often assessed by quantifying the accumulation of BFA-induced endomembrane compartments [36] , [37] . BFA inhibits the activity of ARF–GEFs such as GN , preventing recycling of proteins back to the PM , resulting in their depletion from the PM and accumulation in endomembrane compartments . We treated stem segments with 50 µM BFA for 3 h , but we observed that this treatment had no significant effect on the amount of PIN1 on the basal PM ( Figure 5A ) , and relatively few PIN1 containing compartments were identified . Only 9 of 29 cells examined with optical sectioning throughout the z-axis contained a compartment . This contrasts to results previously described for PIN1 in roots , where after 90 min treatment with 25 µM BFA , the mean number of BFA compartments per cell was more than 1 [37] . These results suggest that either PIN1 endocytosis in stems is BFA-sensitive ( for tissue-dependent BFA effects , see [38] ) and/or PIN1 cycles only slowly in stem segments . To assess the rate of PIN1 allocation to the PM , we used fluorescence recovery after photobleaching . In root cells , after photobleaching total PIN1 signal from a cell , nonpolar PM PIN1 was detected after 100 min [39] . We bleached only the basal PM PIN1 of xylem parenchyma cells , and no significant fluorescence recovery was detected 90 min after bleaching , with little visible effect even after 3 h ( Figure 5B ) , suggesting low insertion rates for PIN1 from either intracellular stores or de novo synthesis . This suggests that at steady state , either cycling rates in stems are low or the fraction of PIN1 in intracellular compartments is very low . To test whether GR24-triggered PIN1 depletion is clathrin-dependent , we determined the effect of the clathrin inhibitor , A23 [40] . A23 treatment alone had no effect on PIN1 levels , providing further evidence for a low rate of insertion of PIN1 or a low intracellular fraction . However , in the presence of A23 , including a 30-min pretreatment , the ability of GR24 to deplete PM PIN1 was abolished ( Figure 5C ) , whereas when treated with the structurally related but inactive control A51 , GR24 triggered PIN1 depletion from the PM as previously observed ( Figure 5C ) . These results suggest that a rapid , nontranscriptional mode of action of strigolactone is to promote a clathrin-mediated step in PIN1 depletion . Indeed in this experiment , statistically significant depletion of basally localised PIN1 was observed within 10 min . As described above , in roots there is rapid constitutive cycling of PIN1 between the PM and the endomembrane system . However , this cycling is not specific but rather reflects general cycling of many proteins . Treatments that affect PIN1 levels at the PM , such as auxin , BFA , and A23 treatment , also affect many other membrane proteins , such as water channel proteins of the PIP1 and PIP2 families [28] , [36] , [37] . Thus in the root , a major contributor to PIN1 behaviour is general trafficking activity . We therefore tested the specificity of the effects of GR24 on PIN1 PM levels in shoots by assessing its effects on PIP1 [41] . PIP1 levels on the PM were less stable over time than PIN1 levels and were halved after 90 min , regardless of the presence or absence of GR24 ( Figure 5D ) , indicating that GR24 has no effect on PIP1 levels . These results suggest that the effect of strigolactone on PIN1 PM levels in stems is more specific than known mechanisms regulating PM PIN1 levels in roots . The apparent specificity of strigolactone effects on PM PIN1 in shoots raises interesting questions concerning the effect of strigolactones on roots . Various aspects of root development , such as primary root length , lateral root development , and root hair elongation , have recently been shown to be modulated by strigolactones [42]–[44] . The role of auxin transport in these phenotypes is unclear , but there is some evidence to suggest that they are at least in part mediated by differences in auxin transport , either locally in the root or systemically from the shoot . To investigate the relationship between auxin transport and the effects of GR24 on roots , we grew Arabidopsis seedlings for 3 d on agar medium without exogenous hormones , preincubated them for 24 h with various concentrations of GR24 , and then observed their elongation over the next 24 h . In the wild-type , two responses were found: agravitropic root growth and root growth inhibition ( Figure 6; also see [42] ) . With respect to effective doses , agravitropic root growth required very high concentrations of GR24 ( greater than 10 µM , Figure 6A ) , and there was no significant difference between wild-type and max2 mutants in response to 100 µM GR24 ( Figure 6B ) . The very high levels of GR24 needed for this effect and lack of dependence on MAX2 led us to conclude that it is of limited physiological relevance . In contrast , as previously shown [42] , root elongation was inhibited by more physiologically relevant levels of GR24 , with GR24 levels between 3 and 30 µM having a significantly weaker effect on max2 than on wild-type , although dose-dependent inhibition was observed in both these genotypes ( Figure 6C ) . Thus , root growth inhibition by GR24 is partially MAX2-dependent . To assess the involvement of auxin in GR24-induced root growth inhibition , we measured root growth in seedlings treated with 0 . 1 µM 2 , 4-D , a synthetic auxin , or 5 µM GR24 , comparing wild-type , max2 , gn , tir3 , and the auxin signalling mutants axr1 and tir1 ( Figure 6D ) [45] , [46] . The max2 mutant responded to 2 , 4-D as wild-type and showed resistance to GR24; axr1 and tir1 showed resistance to 2 , 4-D and responded normally to GR24; gn and tir3 responded normally to 2 , 4-D but showed mild hypersensitivity to GR24 . Thus , as in the shoot , there is an interaction between GR24 and GN/TIR3 . To test the effects of GR24-treatment and gn and tir3 mutation on PIN1 protein levels in roots , we observed the root tips of 4-d-old Arabidopsis seedlings harbouring a PIN1:PIN1–GFP transgene in either the wild-type , gn , or tir3 genetic backgrounds after a 12-h incubation with or without 10 µM GR24 . Neither GR24-treatment , gn , nor tir3 altered total signal levels or obvious subcellular localisation of PIN1 protein in the root tip ( Figure 6E , F ) . Even after a 48-h incubation , 10 µM GR24 did not alter total signal levels or obvious subcellular localisation in wild-type ( Figure 6G , H ) . These results are consistent with different PIN1 trafficking dynamics in roots compared to shoots , such that relatively modest increases in strigolactone-triggered PIN1 PM depletion have a much more dramatic effect in the shoot compared to the root .
In the 1930s Thimann and Skoog established that auxin synthesized in active shoot apices is transported down the main stem and inhibits the activity of axillary shoot apices in subtending leaf axils [6] , [7] . However , it was rapidly discovered that auxin acts indirectly to inhibit axillary bud growth , and furthermore there was a fundamental paradox in auxin behaviour . On the one hand , auxin inhibited the activity of axillary buds , but on the other , its synthesis and export from active apices protected them from inhibition by other auxin sources [47] . These classical observations are explicable by the auxin transport canalization based model for shoot branching control . According to this idea , all the meristems in a shoot compete for access to a common auxin transport path down the main stem to the root . Rootward auxin transport from each shoot apex is established by the positive feedback process of auxin transport canalization , the dynamics of which are critically dependent on the strength of the bud as an auxin source , the strength of the stem as an auxin sink , and the dynamics of the positive feedback loop at the centre of the canalization process that connects them . Thus , the auxin transport system in the shoot forms a self-organising network through which all shoot apices communicate by contributing auxin into the system , thereby influencing the ability of other apices to export auxin . This mechanism for shoot branching control is attractive because it explains the classical observations mentioned above and readily supports the integration of both local and systemic factors in balancing growth distribution across the shoot . However , the idea remains controversial , largely due to different ideas about the mechanism of action of another branch-regulating hormone , strigolactone . One hypothesis , generally referred to as the second messenger hypothesis , posits that auxin in the main stem up-regulates the production strigolactone , which moves into the axillary buds and inhibits their growth by locally up-regulating transcription of the TCP family transcription factor BRC1 , which is known to be required for stable bud inactivation [14] , [25] , [26] . A second hypothesis assumes that axillary bud activity is regulated by the auxin transport canalization-based mechanism described above and that strigolactone acts by modulating auxin transporter accumulation , thereby modulating the ease with which axillary buds can establish active auxin transport into the main stem ( Figure 7 ) [17] , [19] , [20] , [23] . Thus the mechanism of strigolactone action and the mechanism of auxin-mediated bud inhibition are tightly intertwined , representing two different scenarios for the systemic coordination of growth across the shoot system . The results presented here strongly support the second hypothesis . A particularly striking illustration of this is the ability of strigolactone to promote shoot branching in the tir3 mutant background ( Figure 3 ) , which is difficult to explain if strigolactones act as direct inhibitors of bud growth but is a prediction of the model in which strigolactones act to modulate auxin transport ( Figure 2 ) . It should be noted that the two models can easily be reconciled . For example , the primary mode of action for strigolactone could be on PIN1 accumulation , and the resulting effects on auxin transport could in turn influence BRC1 transcript levels . Up-regulation of BRC1 by strigolactone addition to pea buds has been shown to be independent of new translation , but so far it has only been measured after 6 h [26] , [48] , and no such up-regulation was detected in a similar experiment in rice after 3 h of treatment [49] . In contrast , in Arabidopsis stems , an effect on PIN1 accumulation was observed within 10–40 min of strigolactone application ( Figures 4 and 5 ) , and this effect is also independent of new protein synthesis . It is therefore possible that BRC1 transcript changes are downstream of changes in PIN1 accumulation , and the role of BRC1 could be to stabilise bud inactivation caused by low auxin export . Some stabilizing system to maintain bud inactivity seems intuitively important , because bud activation by the positive feedback inherent in canalization is highly likely to be triggered by stochastic variation in the system . These two models differ in that in the canalization model , strigolactones act systemically on the auxin transport network , including in the bud ( Figure 7 ) , whereas in the second messenger model , they act locally and specifically in buds . The systemic expression of MAX2 in xylem-associated cells and the effect of strigolactone on PIN1 accumulation in the main stem are consistent with systemic action . This mode of action allows strigolactones to modulate bud–bud competition systemically , for example in response to nutrient deprivation [13] . In this context , systemic strigolactone levels determine how many buds can activate , but they do not determine which buds activate . This can be regulated by local factors such as light levels [50] . Thus , both local and systemic modifications to the auxin transport network can integrate different environmental inputs to direct resource allocation across the plant body . A more direct mode of action for strigolactone locally in buds does not have this interesting property . However , the two models , and indeed others , are mutually compatible and could operate in parallel with either species-specific and/or environment-specific variation in their relative importance . Little is known about the molecular mechanism of strigolactone action . Only two genes have been implicated in strigolactone signalling . These are MAX2 , which encodes an F-box protein presumed to be required for the strigolactone-regulated ubiquitination of one or more specific target proteins , and D14 , which encodes an α/β hydrolase protein that binds GR24 , confers signal specificity to the pathway [51]–[53] , and could either act as a receptor or could process strigolactones to form a final bioactive product . The immediate downstream effectors of the pathway are unknown , but the largely nuclear localization of MAX2 [54] and the rapid changes in transcription induced by many F-box-protein–mediated plant hormone signalling pathways [55] have led to an assumption that the primary targets for the strigolactone pathway are also transcriptional . The evidence to support this mode of action is currently quite weak . Few reliable transcriptional readouts for strigolactone response have been identified . These tend to have slow induction kinetics , in the order of several hours , and relatively small fold inductions [26] , suggesting that they may be secondary responses or limited to a small proportion of cells . Microarray analysis of Arabidopsis seedlings treated with or without 1 µM GR24 for 90 min shows that 76% of all the GR24-repressible genes are categorised as auxin-inducible [56] , and thus these transcriptional effects may be indirectly mediated via changes in auxin distribution . Consistent with this idea , we have shown that a rapid translation-independent response to stigolactone addition is changes in PM PIN1 accumulation ( Figure 4 ) . Thus , at least one immediate early target downstream of MAX2 in the stem is not transcriptional but involves PIN1 depletion from the PM by an A23-sensitive mechanism , such as clathrin-mediated endocytosis [40] . The mechanism by which the substantially nuclear MAX2 influences PM PIN1 is not known . However , our data suggest that it is both quantitatively and qualitatively different from the major PIN1-regulatory systems operating in the root . Several lines of evidence support this conclusion . First , in stems , strigolactone response is independent of TIR3 activity , which has been reported to be required for auxin-induced inhibition of clathrin-mediated endocytosis in roots [30] . Second , the effect of strigolactone on PIN1 depletion from the PM in stems appears to be more specific than the systems operating in roots , since it does not affect the PM levels of PIP1 , although we have not excluded targets beyond PIN1 . Third , the MAX2-dependent effects of strigolactones on root phenotype are generally less dramatic than those observed in shoots , both with respect to cell biological and whole organ-level phenotypes . Although more modest than the effects on shoots , long-term effects of GR24 treatment on PIN1 accumulation in the root tip have been detected following 6 d of growth in the presence of 5 µM GR24 [44] . These effects have been correlated with reduced shoot-to-root auxin transport , suggesting that they represent a transcriptional response to low auxin rather than the protein trafficking mechanism we propose here . Consistent with this idea , the accumulation of multiple PIN proteins is affected in these root tips , including in cell layers where MAX2 is not expressed at detectable levels [44] , [54] . However , although we found only weak MAX2-dependent root growth inhibition by GR24 , this occurred with equal effect in the auxin signalling mutants , axr1 and tir1 , suggesting that GR24 reduces root growth at least to some extent independently of auxin concentration-mediated effects . Similarly , in the trafficking mutants , gn and tir3 , GR24 reduced root growth more severely than in wild-type , suggesting some overlap in the mechanism underlying the control of shoot branching by strigolactone and its effects on root growth . Computer simulations of shoot growth using our canalization-based model consistently reproduce biological results when strigolactone action is ascribed to a linear process of PIN removal from the PM , independent of PIN insertion and auxin flux . Consistent with this idea , bioimaging of PIN1 protein in inflorescence stems revealed a substantial increase in PIN1 protein in the basal PM in strigolactone mutants . Furthermore , GR24 promoted rapid , translation-independent , MAX2-dependent depletion of PIN1 from the PM through a mechanism sensitive to A23 , an inhibitor of clathrin-mediated membrane trafficking . These results are consistent with the hypothesis that strigolactone functions to promote endocytosis of PIN1 from the PM . It is interesting that the phenotypes affected by the max mutations and by strigolactone treatment are generally those where auxin transport canalization has been implicated . In the root tip , canalization is not usually considered to play an important role in PIN accumulation , although auxin-induced changes in the lateralisation of PIN1 in the root endodermis have been described and compared to canalization processes [22] . If the effects of strigolactones on auxin transport are specifically to modulate canalization , then they provide an opportunity to understand better this enigmatic and poorly understood process , which nonetheless provides powerful explanations for complex patterning events in plants and for their impressive developmental plasticity .
All lines are in the Col-0 background . Experiments involving max2 , used max2-3 [57] , max4 , max4-1 [15] , axr1 , axr1-3 [58] , tir1 , tir1-1 [46] , gn , gnomB/E [59] , and tir3 , tir3-101 [60] . Because we found that the tir3-101 line from a public stock had an additional glabrous mutation besides a C-to-T nonsense mutation at the 3 , 095th codon of TIR3 , a tir3-101 line free from the additional mutation was made and used . For bioimaging , each line homozygous for the PIN1:PIN1–GFP transgene cassette [61] was used . For the PIP1 experiments , the UBQ10:PIP1–YFP ( Wave138Y ) fusion line was used [41] . On-soil and axenic growth conditions were as described previously [17] . For quantifying root growth inhibition and agravitropic root growth , axenic seedlings grown vertically for 3 d on hormone-free agar medium were preincubated for 24 h on vertically placed agar medium containing either only the vehicle , GR24 , or 2 , 4-D . For evaluating the root growth inhibition , the root tip position was recorded just after the preincubation and 24 h after; thus , the length of the primary root grown for the 24 h was obtained . For evaluating the agravitropic root growth , preincubated seedlings were placed horizontally; the root tip position was recorded just after the gravistimulation and every hour up to the next 24 h; thus , the index Curvature , which we defined as the change in root tip angle per the length of grown root within a range between 1 and 3 mm , was calculated . Other physiological experiments were as described previously [17] . All simulations were according to the model of Prusinkiewicz et al . ( 2009 ) [20] . For simulating the auxin transport assay ( Figure S2 ) , the two most basal metamers in the main stem of the whole plant simulated for 2 , 000 time steps were used; of these two metamers , the top one provided an initial value of the PIN concentration at the basal face , and the bottom one provided an initial value of the PIN concentration at the apical face . Auxin concentration in the top metamer was assumed to be 10 and constant over time; auxin concentration in the bottom metamer was assumed to be zero initially and change over time according to the Equations 1 and 2 of Prusinkiewicz et al . ( 2009 ) [20] . The auxin concentration of the bottom metamer at time step 10 was calculated from those two initial values of the PIN concentrations and converted to the percentage of wild-type . This percentage is shown as the simulated polar auxin transport level . For imaging PIN1–GFP in inflorescence stems , the most basal part of the primary inflorescence stem of 6-wk-old soil-grown plants was longitudinally halved by hand with a razor blade . The cut surface was immediately observed using light microscopy and a Zeiss LSM 710 confocal microscope to identify xylem parenchyma cells according to both the relative position to xylem vessels and the morphology of the cell . With excitation at 488 nm , images containing emission spectra from 490 to 655 nm were then acquired within a single dynamic range . Reference spectra of GFP and chloroplast autofluorescence were obtained using a PIN1:PIN1–GFP line in the wild-type background and were used for linear unmixing of the images . For its quantitative analysis , only xylem parenchyma cells that appeared intact and were exposed to the cut surface were taken into account , and the intensity of their unmixed GFP signal was measured in a region of the basal PM that was manually traced . Data were obtained in the same way for real-time monitoring experiments , except that sections were observed with Zeiss LSM 780 confocal microscope . With excitation at 488 nm , images containing emission spectra from 507–550 nm and 593–719 nm were acquired simultaneously in separate channels . Data were obtained in the same way for the PIP1 experiment , except excitation was at 514 nm , and emission spectra were acquired from 518–621 nm and 647–721 nm . For photobleaching experiments , a region of interest ( basal PM of xylem parenchyma cell ) was selected and bleached using the 488 nm laser at 50% power for 75 iterations . In all experiments , cells from three or more plants were included for each genotype/treatment , and the results presented are typical of at least two independent experiments . For imaging PIN1–GFP in the root tip , 3- to 5-d-old seedlings incubated for 12 or 48 h on agar medium containing the vehicle or 10 µM GR24 were immersed in 10 µg/ml propidium iodide for 10 min . The primary root was then observed with Zeiss LSM 510 Meta confocal microscope . The GFP signal excited with a 488 nm laser and the propidium iodide signal excited with a 543 nm laser were collected with a 505–550 nm bandpass filter . For its quantitative analysis , the average intensity of the GFP signal was measured in the stele region of each root . Based on the assumption that the root angle after gravistimulation and the number of branches do not always follow the normal distribution , nonparametric methods of Wilcoxon , Steel–Dwass , and Shirley–Williams were used . Otherwise parametric methods of Student , Tukey , Dunnett , and Williams were used . Unless otherwise stated , statistical results of two-tailed tests are shown in graphs in the conventional manner . In Steel–Dwass' and Tukey's tests , different letters denote significant differences at p<0 . 05 . In other tests , no marks or n . s . indicate not significant , and significant differences are indicated by asterisks as follows: p>0 . 05; * p<0 . 05; ** p<0 . 01; *** p<0 . 001 . | Plants can adapt their form to suit the environment in which they are growing . For example , genetically identical plants can develop as a single unbranched stem or as a highly ramified bush . This broad developmental potential is possible because the shoot system is produced continuously by growing tips , known as shoot meristems . Meristems produce the stem and leaves of a shoot , and at the base of each leaf , a new meristem is formed . This meristem can remain dormant as a small bud or activate to produce a branch . Thus , the shoot system is a community of shoot meristems , the combined activity and inactivity of which shape shoot form . Here we provide evidence that growth is balanced across the Arabidopsis shoot system by competition between the shoot meristems . This competition is likely mediated by the requirement of meristems to export the plant hormone auxin in order to activate bud outgrowth . In our model , auxin in the main stem , exported from active branches , can prevent auxin export by dormant buds , thus preventing their activation . Our findings show that a second hormone , strigolactone , increases the level of competition between branches by making auxin export harder to establish . Together , these hormones balance growth across the shoot system , adjusting it according to the environmental conditions in which a plant is growing . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"developmental",
"biology",
"plant",
"science",
"plant",
"biology",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2013 | Strigolactone Can Promote or Inhibit Shoot Branching by Triggering Rapid Depletion of the Auxin Efflux Protein PIN1 from the Plasma Membrane |
African trypanosomes thrive in the bloodstream and tissue spaces of a wide range of mammalian hosts . Infections of cattle cause an enormous socio-economic burden in sub-Saharan Africa . A hallmark of the trypanosome lifestyle is the flagellate’s incessant motion . This work details the cell motility behavior of the four livestock-parasites Trypanosoma vivax , T . brucei , T . evansi and T . congolense . The trypanosomes feature distinct swimming patterns , speeds and flagellar wave frequencies , although the basic mechanism of flagellar propulsion is conserved , as is shown by extended single flagellar beat analyses . Three-dimensional analyses of the trypanosomes expose a high degree of dynamic pleomorphism , typified by the ‘cellular waveform’ . This is a product of the flagellar oscillation , the chirality of the flagellum attachment and the stiffness of the trypanosome cell body . The waveforms are characteristic for each trypanosome species and are influenced by changes of the microenvironment , such as differences in viscosity and the presence of confining obstacles . The distinct cellular waveforms may be reflective of the actual anatomical niches the parasites populate within their mammalian host . T . vivax displays waveforms optimally aligned to the topology of the bloodstream , while the two subspecies T . brucei and T . evansi feature distinct cellular waveforms , both additionally adapted to motion in more confined environments such as tissue spaces . T . congolense reveals a small and stiff waveform , which makes these parasites weak swimmers and destined for cell adherence in low flow areas of the circulation . Thus , our experiments show that the differential dissemination and annidation of trypanosomes in their mammalian hosts may depend on the distinct swimming capabilities of the parasites .
Trypanosomes are extracellular parasites with an exceptionally broad host range [1] . These flagellates thrive in all vertebrate classes and cause severe diseases in man and livestock . Human African trypanosomiasis ( HAT ) , commonly known as sleeping sickness , is a devastating neglected disease of poverty , and trypanosome infestations of livestock cause additional massive economic burden in sub-Saharan Africa . The animal African trypanosomiases ( AAT ) comprise a set of veterinary diseases , of which the cattle sickness nagana and the equine plague surra are the most prominent . Trypanosoma vivax and T . congolense are the nagana pathogens of cattle , but can also cause disease in other mammals , including sheep , goats , pigs , horses , camels and even dogs . Both species have additionally been identified in a wide range of wild animals , including ruminants and suids , but also lions or hyaenas [2] . T . brucei is pathogenic to camels , horses and dogs , but is also prevalent in sheep , goats , cattle and pigs as well as in a wide variety of wildlife species . The broad host range is shared by the human sleeping sickness parasite T . b . rhodesiense in east and southern Africa . T . b . gambiense causes HAT in west and central Africa and has been reported only in pigs and some wildlife hosts [3] . Most African trypanosomes are transmitted by the tsetse fly . Due to recent partial loss of the mitochondrial DNA , T . evansi , now recognised as another subspecies of T . brucei , is no longer capable of infecting the tsetse fly [4] . Consequently , the prevalence of T . evansi is no longer restricted to the sub-Saharan tsetse belt . In fact , mechanically transmitted T . evansi parasites cause surra in horses , mules and cattle not only in Africa , but also throughout large parts of Asia and South America , where the trypanosomes are also found in wild reservoir hosts [5] . Likewise , T . vivax can be transmitted mechanically and hence , has extended its geographic distribution to South America . Thus , many trypanosome species are contagious for a wide range of diverse mammals . This distinguishes them from other important parasites , such as Plasmodium , which infects only a single genus or even species . Toxoplasma infects a wide range of animals , sexual development and oocyte formation , however , occurs only in feline hosts . While those pathogens invade host cells , African trypanosomes prosper extracellularly in the circulation and various tissues . The question arises whether the extraordinary expansion of host range has evolved as a consequence of the extracellular lifestyle . In fact , all AAT-causing trypanosomes face similar challenges of the mammalian immune system . The defence against host immunity is primarily mediated by sequential expression of antigenically distinct glycosylphosphatidylinositol ( GPI ) -anchored variable surface glycoprotein ( VSG ) [6–8] , a feature that is known as antigenic variation . The parasites have been shown to exhibit high rates of membrane trafficking [9 , 10] , which enables internalisation of antibody-VSG complexes on the parasites’ surface [11] . Endocytosis in African trypanosomes is localised to the posterior part of the cell , where membrane exchange occurs solely at the flagellar pocket ( FP ) , a specialised flask-shaped invagination . The rate of endocytosis in mammalian stage trypanosomes is exceptionally high [9 , 10] , and has been implicated in survival and parasite infectivity [12 , 13] . Moreover , membrane transport has been shown to be developmentally regulated [10 , 14 , 15] and endocytosis occurs exclusively via clathrin [16–18] . Trypanosome motility has been linked to the parasites´ survival in the mammalian host [11] , as well as to successful cell division and development [19–21] . Host antibodies targeted to the VSG coat are shunted to the FP by hydrodynamic drag forces due to the directional motion of the parasite [11] . Significant progress has been made in understanding the basic properties of flagellar dynamics and cellular motility of T . brucei in culture media [22–24] . However , the physicochemical conditions of culture media differ from that of mammalian blood , their natural environment . In fact , we had previously shown that culture forms of T . brucei modulate the beat direction of their flagellum in response to purely biomechanical cues [24] . Thus , the microenvironment determines in which direction the trypanosome flagellum beats and hence , in which direction the cell moves . While these results were conclusive for T . brucei laboratory strains , the relevance to natural isolates or other trypanosome species has not been shown . Indeed , we lack any quantitative data on trypanosome motility in their natural host environment . Therefore , by using high spatiotemporal resolution microscopy we measured the swimming behaviour of the salivarian trypanosomes in blood from different hosts . High-resolution microscopy allowed the comparative analysis of trypanosome motility at the cellular level , i . e . flagellum-driven motility was detailed to reveal flagellar beat frequency , wavelength and beat-to-beat velocity . In addition , trypanosome swimming in the host’s bloodstream was simulated using two different methods; the first one using methylcellulose to change the fluid environment´s viscosity and the second using polydimethyl siloxane ( PDMS ) -pillar arrays to mimic swimming between blood cells . Our findings document marked differences in motility patterns and swimming speeds , which are not only species-specific , but may also vary between strains and differentially depend on the micro-environment . We suggest that the type of motion behaviour may contribute to the dissemination and annidation of trypanosomes in their mammalian host .
We compared the swimming behaviour of the animal parasites T . vivax , T . brucei , T . evansi and T . congolense in fresh wet blood films . Trypanosomes were classified by the swimming pattern they exhibit on time scales of seconds to tens of seconds . During these periods , the cells can swim persistently in one direction and cross spaces of several hundred micrometres , but they can also exhibit periods of non-directed movement , in which they do not move much further than their body length , i . e . around 20 microns . This behaviour is reminiscent of the bacterial “run and tumble” motion and trypanosomes have thus been classified as persistent or tumbling swimmers . We have termed cells alternating between the two modes intermediate or switching swimmers ( Fig 1A ) [22 , 23] . The motion patterns show a clear difference between T . congolense on one hand , with 78% to 92% tumbling cells ( n ≥ 300 ) , and T . vivax , T . brucei and T . evansi on the other hand . Interestingly , we also observed differences between strains of the same species . T . vivax strain IL 1392 harvested from infected mice revealed just 23% tumbling cells , while strain IL 2136 showed 63% tumblers in mouse blood ( Fig 1B and 1C ) . When T . vivax IL 1392 was isolated from sheep , 65% of cells were persistent swimmers , while only 18% of parasites were persistently moving when grown in rat ( Fig 1B ) . T . evansi grown in mice also revealed clear strain-specific differences ( Fig 1C ) . In view of the remarkable variability , we devised a series of experiments to unravel the distinguishing basis of the different motion pattern of trypanosomes , both on the population and the single cell level . First , we directly compared the swimming speeds of the different parasites . High-resolution microscopy allows to measure translocations of single cells with 2 ms accuracy [24] . This temporal resolution allowed us to separate the translational swimming phases from even brief tumbling phases . The measurements yielded maximum and minimum velocities , and also the average swimming speeds for tumbling , intermediate and persistent subpopulations ( n = 100 each ) ( Fig 2 ) . For all trypanosomes analysed , the mean speeds of persistent swimmers were at least twice as high as those of tumbling cells . Obviously , the actual swimming stretches of tumblers stay in the defined 25 μm region ( Fig 1 ) . Nevertheless , the maximum speeds measured in these short time periods can reach those of persistently swimming cells . The trypanosomes can reach these high velocities and halt again so rapidly , because their physical environment is characterised by extremely low Reynolds numbers and inertia is irrelevant . Therefore , all accelerating or decelerating forces produce immediate reactions of the cells [25] . The speeds of the intermediates were in between , showing that the average swimming speed correlates with the overall classification of motility patterns . Although the mechanism of translocation allows many cells to reach high swimming speeds , at least on a short time scale , the lower average speeds indicate that many cells do not accelerate equivalently . These results can only be explained by the detailed analysis of the swimming mechanism in the millisecond timescale . A remarkable result was the maximum swimming speed of a subset of persistent T . vivax cells of over 100 μm/s ( Fig 2 ) . These cells represented about 1% of the population and were characterised by a very straight appearance and slim swimming trajectory ( see below ) . In the next step , we compared the effect of whole blood from different mammals on the motion of the parasites . The trypanosomes were incubated in blood freshly drawn from rat , rabbit and cow ( Fig 3 ) . These experiments were done under controlled conditions with parasites from the same mouse infection . In this way , we were able to observe the immediate impact of different blood sources on trypanosome swimming . Fig 3 reveals representative examples of tumbling , intermediate and persistent swimmers in blood from different sources , including the maximum speeds recorded ( n = 20 ) . The microscopic observation of the parasites and the measurement of swimming speeds did not reveal any obvious differences in the motion behaviour of the parasites in blood from diverse sources . Having measured the swimming capacity on the population level , we subsequently proceeded to single cell analyses of motility and morphology to elucidate the potential structural basis for the different swimming behaviours . High-speed analyses on the single cell level and with single flagellar beat accuracy documented the immediate dependence of cell locomotion on flagellar oscillation ( Fig 4 ) . Persistently forward moving cells of all four species exhibited successive flagellar beats originating at the anterior tip of the free flagellum and generating a tip-to-base wave running along the cell body to the base at the flagellar pocket ( exemplified in the annotated S7 Video of the T . vivax cell in Fig 4A ) . Each persistently swimming cell exhibited tip-to-base beating with a fairly constant frequency . Interestingly , cells with the same beat frequency were observed to swim with significantly different speeds . In the case of the two T . vivax strains analysed , about 1% of the population exhibited a slim waveform , i . e . a markedly straighter and thinner appearance , as revealed by the microscopic observation of several thousand parasites . The slim trypanosomes were characterised by the flagellum forming just one wavelength along the entire length axis of the parasite . Thus , the flagellar wave dominates the shape of the cell , thereby producing very effective persistent swimmers that reached high swimming speeds of over 100 μm/s ( Fig 2; single cell analysis in Fig 4A ) . The majority of T . vivax parasites , however , showed a more flexible form of the flagellar wave , resulting in average speeds of around 30–40 μm/s ( single cell analysis in Fig 4B ) . In both conformations the flagella beat with almost identical average frequencies . The slim single-wave type of T . vivax represented an ideal oscillatory conformation of the complete cell , resulting in a highly propulsive travelling wave along the cell body . As in trypanosomes the flagellum is uniquely attached to the cell body , the flagellar oscillation continuously modulates the shape of the cell , rendering it virtually impossible to deduce a static morphological picture of the swimming parasite . We therefore suggest the term “cellular waveform” , which is the product of the flagellar wave , the chirality of the flagellum attachment and the degree of cell stiffness . The overall appearance of the cellular waveform is characteristic for each trypanosome species , but can be modulated in a strain-specific manner . It is to be noted , that the T . vivax slim waveforms readily convert briefly into the normal waveform , when hindered in free movement ( S17 Video ) , whereas the reverse change of the unobstructed normal waveform converting into the slim waveform has not been observed . In fact , the slim waveform is not necessarily longer or slimmer than other trypanosome cells but only appears so in motion , because of its straight cellular waveform and high speed . This fact , together with the occurrence before peak parasitaemia , argue against this form being the late stationary developmental stage identified in infected mice , which additionally has a sub-terminally located kinetoplast [27] . Thus , two different subpopulations of T . vivax are present during infections , which do not represent distinct morphotypes but different cellular waveforms , once more underlining that motion and morphology are interlaced in trypanosomes . T . brucei and T . evansi reveal comparable swimming patterns , velocities and flagellar beat frequencies ( Figs 1C , 2 , 4D and 4E ) . This is expected , as T . evansi most likely represents a mutant of T . brucei , adapted to mechanical transmission . Interestingly , however , the cellular waveform distinguishes both parasites . Fig 4E shows a typical example of a persistent forward swimming T . evansi cell . The parasites revealed a more flexible appearance with several short wavelengths of the flagellum running along a rather elastic cell body . This produced the characteristic ‘curly’ waveforms of T . evansi and the resulting smooth , snake-like forward movement , which was observed to be strongly influenced by mechanical interaction with the environment ( S12 Video , see single beat analysis below ) . The proportion of persistent swimmers in T . congolense populations was comparatively low . Nevertheless , the cells were able to swim forward , albeit for shorter periods , with speeds around and significantly above 20 μm/s ( Fig 4F and 4G ) . The oscillation of the anterior flagellar tip was not as pronounced as in other species , as there is no free part of the flagellum . Thus , the beating resulted in weaker oscillations and altogether a less propulsive cellular waveform of the short rigid cell . Interestingly , T . vivax IL 1392 and T . congolense KETRI 3827 showed a marked reduction of swimming speed in sheep blood compared to mouse blood ( Fig 4H and 4I ) . This was not due to any adherent interaction with the blood cells ( S15 and S16 Videos , see [28] ) . We have specifically analysed freely swimming parasites in this work . Thus , the host environment influences the swimming performance of the two most different trypanosome swimmers in a comparable way . In order to better understand the cellular waveform as the basis for the motile behaviour of different trypanosome species and strains , we analysed some morphological key-parameters , such as the flagellar attachment , the cell dimensions and the apparent cell flexibility . We fluorescently labelled the cell surface of T . vivax , T . brucei , T . evansi and T . congolense and recorded a several hundred three-dimensional microscopic image-stacks , in order to produce three-dimensional representations of the cells ( Fig 5 ) . In all trypanosomes analysed the flagellum was closely attached to the cell body and did not constitute part of an elastic undulating membrane . The two T . vivax strains showed a rather rigid conformation of the cell body . The flagellum attaches in a right-handed path around the cell . The turn reaches around 180° , but the path is less curved as in T . brucei , possibly due to a greater stiffness of the cell body ( Fig 5A and 5B ) . The morphology of the cell culture-adapted T . brucei strain MITat 1 . 6 had previously been described as an on average s-shaped cell body with an attached flagellum running from the flagellar pocket in a left hand 180° turn around the cell and continuing along a thinning body to the very flexible anterior end [24] . In comparison , we found that the cells of the T . brucei ILTat 1 . 4 strain analysed here , generally were larger and showed a longer and thinner appearance . Interestingly , the flagellum revealed mirrored chirality , i . e . it followed a right-handed turn of 180° around the cell body , from the emergence at the flagellar pocket to the free anterior end ( Fig 5C ) . Analysis of the pleomorphic T . brucei AnTat 1 . 1 strain then showed the same right-handed chirality ( Fig 5D ) . We conclude that in T . brucei the flagellum generally follows a 180° path around the cell and contributes to the characteristic cellular waveform , which however , is not influenced by the rotational direction of flagellum attachment . The cell body of T . evansi represents the most flexible of the cell types analysed here . The flagellum followed a full 360° right-handed path around the cell body from posterior to anterior ( Fig 5E ) . These features give T . evansi cells their typically curled appearance , which distinguishes them from T . brucei and contributes to their characteristic cellular waveform . T . congolense cells were typically the smallest and the cell body seemed the least flexible ( Fig 5F and 5G ) . The flagellum followed a relatively straight path along the cell body , the flagellar turn still producing asymmetry , but usually far less than the 180° half turns of T . brucei or the full turn T . evansi ( Fig 5C–5E ) . There was hardly any free part of the anterior end of the flagellum visible . This explains the relatively ineffective force generation of the small and stiff cell bodies of T . congolense , making these parasites the slowest swimmers . Interestingly , the stumpy form of T . brucei shares similar characteristics with T . congolense , a reduced cell length and the lack a long free anterior part of the flagellum , although the cellular waveform is more flexible ( Fig 5D , right ) . Like T . congolense , the stumpy cells are slow swimmers . In fact , most cells in a stumpy trypanosome population tumble , and only at higher viscosities the parasites swim for short periods . ( S18 Video ) . In conclusion , we can clearly discern characteristic 3D-morphometric features of different trypanosomes that contribute to their cellular waveform and hence , influence the movement of the parasites . So far , we had analysed regular persistent forward movement and obtained an overview of the various morphotypes and cellular waveforms that could be correlated to the typical swimming performance of different trypanosome species . Using microscopic datasets of sufficient spatiotemporal resolution , we now analysed millisecond variations of swimming patterns , which were observed to happen with or without environmental influences . In view of the variability documented in the above , we asked if the basic swimming pattern ( i . e . forward , backward and tumbling motion ) were produced by the same mechanism on the scale of single flagellar beats in all trypanosome species analysed . Trajectories of swimmers were followed in wet blood films and the cells´ translocation speed was measured for each flagellar beat ( Fig 6 ) , as defined in the analysis shown in Fig 4 . The full trajectories are shown in S8–S16 Videos . These analyses detailed the responses of individual cells to the forces produced by the flagellar waves in the surrounding blood serum , and to the obstacles encountered therein . As shown above , two subpopulations of T . vivax exist in freshly drawn mouse blood , which are characterised by distinct cellular waveforms . The cell in Fig 6A ( S8 Video ) represents the slim waveform ( about 1% of all cells ) and moved persistently forward with an average speed of 95 μm/s and a flagellar beat frequency of 29 Hz ( Fig 6A ) . Notably , when the anterior tip of the flagellum hit an obstacle frontally , as exemplified by a red blood cell , the cell came to an immediate halt ( beat 14 , Fig 6A , S8 Video ) . The beat frequency did not change , i . e . regular flagellar oscillation did not stop , and with the next beat , the cell immediately continued its forward movement with the same speed . At the high spatiotemporal resolution used here , the two-dimensional speed measurements using a single reference point automatically produce fluctuations , depending mainly on the rotation of the trypanosomes and their three-dimensional helical path . Therefore , the single beat measurements ( Fig 6 , orange lines ) are overlaid with averaged velocities ( average of five successive beats , blue lines ) . Both speeds are shown to detect short-term changes in movement that were masked by averaging ( i . e . beat 14 in Fig 6A ) , underlining the importance of high temporal resolution . The second T . vivax example shows the predominant slower normal waveform ( Fig 6B , S9 Video ) . The average speed of this persistent swimmer was 30 μm/s , with a flagellar beat frequency of 27 Hz . The parasite swam through a group of blood cells and came to a brief halt during beats 12–14 . In this case also , the beating frequency did not change , but the flagellar wave was not propagated along the cell body to the posterior end , rendering these tip-to-base beats ineffective . The trypanosome then moved forward again for several beats before it started to produce base-to-tip waves ( beat 22 ) . Flagellum beat reversal has long been observed in trypanosomatids , but its biological relevance is still not clear [29–32 , 24] . In this example , the reversal of the flagellar wave , initiated by base-to-tip beats , caused an immediate reversal of translocation . This reversal led to a tumbling phase of around 800 ms ( beats 22–33 ) , during which the cell produced seven base-to-tip beats . The reverse waves were mostly incomplete and interrupted by short periods without effective force generating flagellar movements ( shown with a frequency value of 0 Hz ) . Therefore there was only a slight backward translocation of the cell . After the tumbling phase , the cell resumed flagellar beating , immediately producing full tip-to-base waves with the same frequency as before the tumbling period , effectively propelling the cell forwards again at average speeds of around 40 μm/s . This brief tumbling phase changed the parasite’s forward swimming direction through the reorientation of the flexible anterior part of the flagellum . The tip-to base waves running along the cell body cause the cell to rotate and break out in a new direction that is targeted by the flagellar tip , in this case a course change of about 30 degrees ( S9 Video ) . The analysis in Fig 6C shows a normal waveform T . vivax IL2136 cell beating persistently with an average frequency of 14 Hz ( S10 Video ) . Despite the relatively constant beat frequency , the average speed fluctuates around a mean of 18 μm/s . This is partly due to mechanical interactions , as at around beat 15 the anterior tip of the flagellum collides with another trypanosome slowing forward movement . On the other hand , around beats 33 and 47 higher speeds are measured , as the flagellum interacts laterally on both sides with red blood cells and other obstacles . Here the resistive force produced as the flagellum wriggles through confined spaces accelerates the cell . The analysis in Fig 6D of a T . brucei cell swimming through relatively dense red blood cells , shows persistent swimming stretches with an average speed of 24 μm/s and a beat frequency of 18 Hz , interrupted by several short stops which are caused by discontinued flagellar beating without obvious mechanical hindrance . Beat 61 and several following are base-to-tip beats with an average frequency of 5 Hz which effectively propel the cell backwards at average speeds of up to 15 μm/s through an obstacle-free area . The single beat analysis of a T . evansi parasite details the impact of the characteristic cellular waveform with its elastic appearance on trypanosome propulsion ( S12 Video ) . In the first part of the trajectory , the cell moved persistently with an average speed of 42 μm/s and a beat frequency of 15 Hz ( Fig 6E ) . During this movement there was one incomplete base-to-tip wave ( beat 8 ) . This caused the cell to decelerate and reoriented the anterior tip of the flagellum with the next tip-to-base beat . With the following tip-to-base waves , the cell accelerated back to higher swimming speeds in the altered direction . In its further path , the trypanosome experienced significant resistance from relatively densely packed erythrocytes . While the cell swam in confinement between blood cells , it retained a relatively constant beat frequency , but was routed in a circular path with reduced swimming speed . Although the general motility pattern of T . congolense differs greatly from that of T . vivax ( Fig 1B and 1C ) , we found that the underlying mechanism of bi-directional flagellar beating is conserved in all trypanosomes analysed . Fig 6F shows two stretches of a T . congolense IL1180 cell swimming persistently forward , interrupted by a backwards swimming phase . Interestingly , in both directions the cell reaches approximately the same speeds and flagellar beat frequencies . Fig 6G exemplifies a T . congolense KETRI 3827 cell that swam forward with an average speed of 17 μm/s and a flagellar beat frequency of 6 Hz . The cell then reversed its beat direction , causing an immediate backward propulsion , accelerated forward again using three tip-to-base beats and then entered a tumbling phase of about 3 seconds , consisting of five base-to-tip beats and millisecond periods without any flagellar movement . After this phase , the cell resumed forward swimming with basically the same frequency and speed as before tumbling . The last example shows a T . congolense KETRI 3827 cell swimming persistently forwards in sheep blood at reduced speeds as compared to this species swimming in mouse blood ( Fig 6H , S15 Video ) The above single cell analyses underlined that the principle of flagellar propulsion is conserved in all studied trypanosome species . The marked differences shown for short- and medium-term swimming behaviours is explained by the dynamic and morphological characteristics of a given cell , i . e . its cellular waveform that is controlled by beat frequency , length and amplitude of the flagellum and the direction of flagellar waves running along the cell body . We have shown that trypanosome motility can be influenced by the direct physical environment , in other words , by the amount and distribution of resistive force on the distorting flagellum and cell body [24] . In the wet blood films used for the characterisation of swimming patterns on the single cell and population level , the cells move in an environment with random concentration and distribution of blood cells . This is different from the circulation , as blood is a self-stirring fluid . Therefore , we used systems that more closely mimic certain aspects of the in vivo environment of bloodstream form trypanosomes . Therefore , we adjusted the viscosity of blood samples by addition of methylcellulose . A 0 . 2% methylcellulose solution has a kinematic viscosity ( 2 . 5 mPa·s ) around 150% higher than that of serum . The addition of methylcellulose to a concentration of 0 . 4% results in a viscosity ( 5 . 2 mPa·s ) comparable to fast flowing normal whole blood ( 4–5 mPa·s ) and the viscosity of 0 . 6% methylcellulose ( 25 mPa·s ) is in the upper range of reported blood viscosity . Values above that ( 0 . 8% methylcellulose ) are in the lower range of oil viscosities at room temperature . For each condition the trajectories of 300 parasites were analysed and scored as in Fig 1A . T . vivax showed a viscosity-dependent decrease of tumbling parasites , from the already comparatively low value of 27% in serum to just 3% in 0 . 4% methylcellulose ( Fig 7A ) . At higher viscosities the numbers of persistent swimmers decreased and the population almost entirely consisted of intermediate swimmers . The average swimming speeds were only marginally faster at blood viscosities ( Fig 7B ) . The additional presence of blood cells reduced the swimming speed by about 20% ( Fig 7C ) . This is in agreement with the swimming behaviour at higher viscosities , as raising the concentrations from 0 . 4% to 0 . 8% methylcellulose , successively reduced the average and maximal speeds of the parasites . This means that T . vivax swimming is optimal at viscosities prevailing in the bloodstream . High viscosity had the opposite effect on the swimming speed of T . brucei cells . The parasites reached maximum average velocities at methylcellulose concentrations of 0 . 4 and above . Even at 35 mPa·s the cells swam with maximal velocities of more than 50 μm/s ( Fig 7B ) . The percentage of tumblers was decreased to 14% at upper blood viscosity ( 25 mPa·s ) and intermediate swimmers dominated the population ( Fig 7A ) . The additional presence of blood cells only marginally affected the swimming performance at higher viscosities ( Fig 7C ) . Thus , T . brucei ILTat1 . 4 motion reached maximum speeds at physiological blood viscosity levels of around 4 to 5 mPa·s and the cells remained comparably motile even at higher viscosities . This was comparable to data published for T . brucei MITat 1 . 6 cells [24] . The motility of T . evansi also increased with rising viscosities . The proportion of tumbling cells decreased from 78% at 1 mPa·s to just 9% in 0 . 6% methylcellulose ( Fig 7A ) . Furthermore , swimming speeds increased with rising viscosity , average speeds rising more than threefold in 0 . 4% methylcellulose . The maximal recorded speed was 40 μm/s , which is four-times higher than in serum ( Fig 7B ) . The presence of blood cells only had little effect on the average persistent speed ( Fig 7C ) . The higher speeds of T . brucei compared to T . evansi correlate with the increase of persistent T . brucei swimmers up to a viscosity of 5 mPa·s . Although the decrease of tumbling cells is comparable in both subspecies , T . evansi cells are virtually all intermediate swimmers , meaning they have a significantly higher frequency of beat reversals and tumbling phases , irrespective of the viscosity of the environment . Another marked difference between T . evansi and T . vivax parasites was the influence of changing viscosity on swimming direction . More than 50% of T . evansi cells swam persistently backwards in serum ( S19 Video ) , whereas after raising the viscosity to low blood levels ( 3 mPa·s ) , backward swimming was no longer observed . T . vivax showed a different behaviour , with 90% of cells showing short-term wave reversals at very high viscosity ( 35 mPa·s ) . So , whereas T . brucei and even more so T . evansi , seemed to be adapted to highly viscous surroundings by increasing their forward swimming persistency and speed , the swimming behaviour of T . vivax is clearly more adapted to high speed and forward persistency in more fluid surroundings . This could constitute an advantage for survival of T . brucei and T . evansi in tissue spaces , whereas T . vivax appears more adapted to viscosities prevailing in blood serum . The influence of viscosity changes on the behaviour of T . congolense was much less pronounced . The amount of tumbling cells was slightly less at 3 mPa·s ( 82% ) than in serum ( 94% ) . Also at higher viscosities the number of swimming cell did not increase ( Fig 7A ) . The maximal speed ( 22 μm/s ) of a T . congolense cell was observed at 3 mPa·s ( Fig 7B ) and the average speed reached 12 μm/s at 25 mPa·s . However , there was no consistent trend of increased swimming performance . Thus , T . congolense is not generally dependent on elevated viscosities for motion . However , the cells are capable of swimming short stretches with speeds over 20 μm/s in environments of elevated viscosity , which is important for antibody clearance . Also T . congolense was able to reverse swimming direction for short stretches ( see example in Fig 6F and 6G ) . These backward swimming paths were also slightly increased in 0 . 2% methylcellulose , but in neither direction did T . congolense cells stand out as persistent swimmers , as the parasites were constantly switching back to tumbling phases . In summary , we show a significant increase in directional speed of T . brucei and T . evansi after raising the environmental viscosity to the range of mammalian blood . T . vivax , on the other hand , shows a decrease in swimming efficiency in higher viscosities and T . congolense seems to be oblivious to viscosity changes . This documents species-specific swimming efficiency profiles in changing physical environments . The use of methylcellulose allowed us to reliably measure the effects of various fluid viscosities on trypanosomes . The macroscopic viscosity of blood , however , is largely due to the presence of blood cells , which produce a non-homogenous system of fluid and obstacles with larger dimensions than those of the microscopic parasites . In order to account for the blood topology , we used arrays of PDMS pillars with diameters in the range of red blood cells , which were differently spaced in order to generate a “frozen suspension” of blood cells . It had previously been shown that the curvature and swimming trajectory of cultivated T . brucei cells are adapted to swimming between such regularly spaced obstacles [24] . Pillar arrays allowed us to measure the maximum performance of swimmers that were moving in a perfectly 2D-horizontal path . In the case of flowing blood , the mean spacing of red blood cells would provide an array of regular obstacles around the trypanosomes in all three dimensions , so the bending flagellum could flexibly use the provided channels in any direction . The cell would always move in an optimal 3D-environment . The diameter of pillars used here was 8 , 10 or 12 μm . Arrays consisted of either same-sized pillars or alternating patterns of mixed pillars with constant spacing . The maximum speed of T . vivax , T . brucei and T . evansi cells was increased in pillar arrays , whereas T . congolense did not show any significant change in motility ( Fig 8 ) . T . vivax reached the highest peak velocity of 205 μm/s in 3 μm spaced pillars of 8 μm diameter . T . brucei showed a peak velocity of 175 μm/s in 4 μm spaced arrays of 10 μm pillars and T . evansi also reached over 100 μm/s in 4 μm spaced pillars with diameters of 8 and 10 μm . These results show that the highest speeds were achieved with a spacing of 4 μm , with T . vivax reaching even higher velocities in narrower spaces of 3 μm . This corresponds well to the calculated average spacing of blood cells in the mammalian bloodstream [24] . Thus , a physiological spacing and pillar diameters of 8–10 μm , which corresponds well to erythrocyte dimensions , promoted the highest parasite swimming speeds . Those high speeds were reached when the cells were able to thread through the pillar arrays by fitting their waveforms . The cells swam with highest directionality when they were ideally aligned in the horizontal plane of the arrays . As soon as they deviated from that plane in z-direction , they left their “pillar channel” , lost the constant resistive force of the pillar walls and consequently , did not reach maximum speeds , even if they continued with persistent forward swimming ( S19–S21 Videos ) . The swimming persistency of T . congolense was not sufficient to enable the cells to take advantage of the pillar arrays . The cells did not exploit the resistive forces provided by the pillars for optimised motion persistency or swimming speed . The T . congolense parasites remained in their predominantly tumbling state . Taken together , for trypanosome species that are sensitive to viscosity changes , we can define environmental conditions that allow cells to reach high directional swimming speeds . We assume these conditions to mimic the environment in the host and thus conclude that these species are adapted to fast persistent swimming in this habitat , whereas T . congolense seems to have adopted a different strategy and does not depend on fast directional swimming in the bloodstream . In a last set of experiments , time-course analyses of antibody removal was conducted in T . vivax , T . brucei and T . congolense ( S1 Fig ) . The trypanosome surface was uniformly covered first with biotin and then with anti-biotin antibodies conjugated to green fluorescent dye at 0°C , when neither movement nor endocytosis take place . After a 20 second incubation at 37°C , most of the surface signal was internalised in all species . Consequently , after 40 seconds of incubation , further internalisation was observed and processing of the signal began as evidenced by the increased number of endosomal vesicles . Sorting and processing of the signal continued during a 5-minute period after uptake of the antibodies and after 10 minutes the signal intensity was reduced in all species , indicating the digestion of antibodies . In summary , all three species investigated were capable of clearing VSG-bound antibodies with comparable rates of endocytosis [11] .
African trypanosomes thrive in diverse host environments ranging from tissues and circulation of many vertebrates to the digestive tract of the tsetse fly . Cell motility appears to be crucial for the completion of the trypanosome life cycles in the fly and vertebrate hosts [19 , 21 , 33 , 34] , and studies investigating motility-dependent mechanisms , like antibody clearance [11] , have led to efforts to elucidate and quantify the exact mechanism of the complex three-dimensional movement of the cells [35] . The results obtained , in turn , led to the concept of how the physical environment could affect the motility of trypanosomes , with implications for the parasites behaviour in the diverse fluids and tissues of their varied hosts [24 , 36] . In order to achieve an understanding of the mechanisms of trypanosome motility and its relevance in natural infections , we aimed at a comparative analysis of the motility of different parasite species . Originally , we observed parasites from naturally infected cattle at Shimba Hills National Park ( South coast , Kenya ) . Although the infection prevalence was high , parasitaemia was expectedly low and thus , did not allow sufficiently quantitative conclusions . Therefore , in the present work , we compared the motion behaviour of different trypanosomes in a series of experiments , ranging from experimental animal infections to single cell analyses in micropillar arrays . Two strains each of T . vivax , T . brucei , T . evansi and T . congolense were chosen for examination . We resolved clear motility patterns , characteristic for each species , however , also found variations between strains of the same species . In a nutshell , T . congolense stands out as a low motility trypanosome . Most cells did not reveal long periods of persistent swimming . T . brucei and T . evansi , on the other hand , as well as T . vivax , showed high proportions of persistent swimmers . T . vivax is capable of persistent and fast swimming with speeds of up to 10 body lengths per second , which is respectable for any swimming organism . For comparison , mammalian sperm can swim with speeds of about 5 body lengths a second . Although T . vivax parasites clearly swam fastest , all species reached peak velocities of more than 20 μm/s , a speed theoretically required to assure cell surface clearance of host-derived antibodies by hydrodynamic drag . Consequently , antibody clearance assays revealed that all species , including T . congolense , removed antibodies from the cell surface with fast kinetics . The T . brucei strains ILTat 1 . 4 and AnTat 1 . 1 presented higher percentages of motile cells when compared to the published data for the monomorphic MITat 1 . 6 strain [24] . This could reflect an adaptation to long-term low viscosity cultivation , underlining the necessity to interpret results from cell culture cautiously . However , we also found clear differences between the motility patterns of the two T . vivax strains analysed , although both were harvested from mice and are not culture-adapted [37] . Interestingly , the host appeared to be less decisive for the trypanosome motion pattern . Although the number of persistent T . vivax swimmers was higher when recovered from sheep compared to the population from mice , the proportion of tumblers was similar . The relatively low influence of host blood on motility patterns was confirmed by directly comparing the swimming performance of the different trypanosome species in blood freshly drawn from various mammals . High spatiotemporal resolution analyses on the population and the single cell level revealed that all trypanosome species are able to move persistently by continuous beating of their attached flagellum , producing flagellar waves running from the anterior tip to the flagellar pocket , located at posterior pole of the cell . Importantly , all cells are capable of producing base-to-tip waves , which generate a force directly opposing the productive tip-to-base waves . This results in the stalling of forward movement , or , given a pause in tip-to-base beating , in backwards motion . The alternating forward and backward movements generate the characteristic tumbling trajectories with minimal net displacement of the parasite . Thus , irrespective of swimming performance , the basic mechanisms of flagellar beating are the same in all salivarian trypanosomes . The computation of three-dimensional surface models of fluorescently labelled trypanosomes , together with extended single-beat analyses of swimming cells unveiled a complex picture of distinctive dynamic parasite morphologies . In African trypanosomes the flagellum is closely attached to the cell body in a rather non-elastic manner . The parasites lack an undulating membrane , which would uncouple the flagellar force from the cell body , as is the case in Trichomonas for example . Owing to its firm attachment , the trypanosome flagellum continuously deforms the whole cell body , and due to frequent beat reversals this deformation is not periodic , which increases the number of forms the cells adopt . Thus , the actual cell shape is a product of the flagellar oscillation ( wavelength and amplitude ) , the chirality of flagellum attachment and the stiffness of the cell body . We have termed this function the cellular waveform . It essentially describes the dynamic pleomorphism of the parasite during a specific life cycle stage . The degree of pleomorphism of a trypanosome species depends on the variability of motion patterns and hence , on the range of cellular waveforms the parasites adopt . Examples of cellular waveforms span from the slim waveform of T . vivax , where the entire cell adopts a shape of one large wavelength , to the compact and stiff waveform of T . congolense . T . vivax trypanosomes are generally large cells with a long free anterior part of the flagellum beating with high frequencies , while T . congolense is characterised by small cell bodies , which dampen the low frequency beating of the completely cell-attached anterior end of the flagellum . In between , we find T . brucei and T . evansi , which possess similar cell shapes and flagellar frequency ranges . These trypanosomes both feature at least two wavelengths of the flagellum following a helical path around the cell body and are rather difficult to distinguish in static pictures . This changes when the cellular waveform is considered , because a curliness of motion trajectories is characteristic for T . evansi . The structural basis for this appearance becomes visible in the 3D model of the parasites; the flagellum follows a complete turn ( 360° ) around the cell body from the flagellar pocket to the anterior tip . This differs from the 180° turn that produces the rotational movements in T . brucei and T . vivax and the more or less straight flagellar path of the stiff T . congolense cells . Flagellum chirality is not necessarily species-specific . T . brucei and T . evansi are regarded as subspecies and show different cellular waveforms . Furthermore , two strains of T . brucei reveal flagellar attachment of mirrored chirality . While the flagella of the ILTat 1 . 4 and AnTat 1 . 1 strains are attached in a 180° right-handed turn , the monomorphic MITat 1 . 2 strain reveals 180° left-handed chirality [24] . Thus , although the chirality of flagellum attachment matters , as it causes the cellular asymmetry required for the rotating motion of the trypanosomes , the rotational direction of flagellum attachment appears irrelevant . The cellular waveform concept is responsive , i . e . it considers the impact of the micro-environment on cell dynamics . Consequently , variations in fluid viscosity and the application of micro-sized obstacles in form of PDMS pillars documented a surprising variability in the motile behaviour between different salivarian species . T . vivax exhibits the highest swimming efficiency in low to medium viscosity surroundings , corresponding to conditions in the peripheral bloodstream , but shows virtually no persistent motility in an environment of higher viscosity . The slim cellular waveform perfectly exploits the spacing between blood cells , resulting in the highest overall trypanosome speed measured in narrow-spaced , erythrocyte-sized pillars . Thus , T . vivax parasites are specialised for fast movement in the densely packed mammalian circulation , however , they are less capable of navigating in tissue spaces . This is compatible to the even distribution of T . vivax in the circulation [37 , 38] . The trypanosomes produce cyclical waves of heavy parasitaemia , to which the host often succumbs , when massive terminal thrombosis occurs in large blood vessels [38–41] . The lack of tissue infiltration in the host is compatible with the weak motility performance of T . vivax in experimental high viscosity environments . T . brucei and T . evansi swim faster and more persistently in fluids with the viscosity of blood and they make use of surrounding obstacles like blood cells to reach maximum swimming speeds . While this suggests an adaptation to the bloodstream , the fact that these species swim even more efficiently in viscosities higher than that of blood indicates that they can navigate in confined surroundings , such as tissue spaces . This correlates well with the known distribution and pathology of the T . brucei group , which are mainly tissue-invading trypanosomes [42–47] . The trypanosomes spread through tissue spaces and the lymphatic system , but are either absent or present in peripheral blood in rather small numbers only [48] . The cellular waveforms of these parasites in fact reflect the capability to manoeuvre in tissues , as the helical attachment of the flagella produces the eponymous auger-like rotating movement , while the cell body flexibility allows for probing narrow spaces in all three dimensions , which is further assisted by the readily triggered backward movements . Therefore , the members of the T . brucei family seem to be adapted to efficiently navigate in the tissues of their hosts . Finally , T . congolense motility behaviour is essentially unperturbed by changing viscosities or the presence of obstacles . Thus , T . congolense behaves fundamentally different compared to other salivarian trypanosomes and might exploit cell motion in another way . The waveform of T . congolense suggests an adaptation to a mainly stationary mode of motility , microscopically apparent as tumbling , regardless of the physical properties of the surroundings . This might explain the distribution of T . congolense within the host: it is strictly a ‘plasma parasite’ , preferentially found in smaller capillaries and venules of organ tissues , like liver , kidney and spleen , where low flow conditions prevail [45 , 49 , 50] . The tendency to adhere via the flagellum to endothelial cells is well described for T . congolense . More vigorous swimming would most certainly interfere with cell attachment [28 , 51] . In tsetse-infested areas , a majority of susceptible animals are co-infected with different trypanosome species . Differential dissemination to the circulation of one species and the simultaneous invasion of tissues by another should greatly contribute to the complex pathogenesis of infection [45] . Our results are compatible with a scenario in which distinct swimming capabilities and cellular waveforms allow the parasites to populate distinct anatomical niches within the same host . As these niches are extracellular , it is tempting to speculate that they do not markedly differ between natural hosts , which in turn could explain the comparable dissemination of a given trypanosome species in a wide range of different mammals . Even if we are not yet able to simulate trypanosome movement in flowing blood perfectly or measure motility with similar accuracy as achieved here directly in the bloodstream , we now have a reasonable indication of how different species behave in response to varying hydromechanical forces . The next step will be the analysis of cell motility in exactly defined fluids and materials in vitro , using microfluidic systems and in vivo , exploring living circulation systems compatible with high-resolution microscopy . Only in controlled environments like these , systematic mutation analyses of trypanosome motion and the underlying cellular waveforms will become feasible .
T . congolense IL 1180 , T . vivax IL 1392 , T . vivax IL 2136 and T . b . brucei ILTat 1 . 4 were obtained from the trypanosome bank at the International Livestock Research Institute ( ILRI , Nairobi , Kenya ) . T . congolense IL1180 is derived from STIB 212 , which was isolated from a lion in the Serengeti area [52] . T . vivax IL1392 is a derivative of Y 486 , originally isolated from a steer in Zaria , Nigeria [53] . T . vivax IL 2136 is derived from IL10-E4 , isolated in Yakada , Nigeria in 1973 . T . b . brucei ILTat 1 . 4 is derived from EATRO 795 , originally isolated from bovine blood in Nyanza , Kenya ( http://tryps . rockefeller . edu/trypsru2_pedigrees . html ) . T . evansi KETRI 2479 , T . evansi KETRI 4009 and T . congolense KETRI 3827 were obtained from Kenya Agricultural and Livestock Research Organization—Biotechnology Research Institute ( KALRO-BRI , Kikuyu , Kenya ) trypanosome bank . T . evansi KETRI 2479 was isolated from a camel in Ngurunit , Kenya in 1980 . T . congolense KETRI 3827 was isolated in Lamu , Kenya in 1997 and T . evansi KETRI 4009 in Marsabit , Kenya in 2010 . T . brucei AnTat 1 . 1 were obtained from the Wuerzburg trypanosome strain collection and are a derivative of LUMP581 , originally isolated from a bushbuck in Mavubwe , Uganda ( http://tryps . rockefeller . edu/trypsru2_pedigrees . html ) . This study was undertaken in adherence to experimental guidelines and procedures approved by the Institutional Animal Care and Use Committee ( IACUC , Ref: C/TR/4/325/125 ) by the Trypanosomiasis Research Centre of the Kenya Agricultural Research Institute ( KARI-TRC ) . These IACUC regulations conformed to national guidelines provided by the Kenya Veterinary Association . Bloodstream forms of T . congolense , T . vivax , T . brucei and T . evansi were inoculated intraperitoneally into Swiss mice and Sprague-Dawley rats . Mice were immunosuppressed before infection ( 300 mg kg-1 cyclophosphamide ) . A drop of blood from the tail was obtained daily to monitor the parasitaemia microscopically . T . vivax IL 1392 and T . congolense KETRI 3827 were used in comparative sheep and mice infections . The sheep were housed at KALRO-BRI , whereas small laboratory animal experiments were conducted at the International Centre of Insect Physiology and Ecology ( icipe , Nairobi , Kenya ) . The trypanosomes were expanded in immunosuppressed donor Swiss mice and harvested at a parasitaemia of 107 cells/ml . The trypanosome containing blood was diluted in PSG buffer ( 0 . 15 M sodium phosphate , 0 . 1 M NaCl and 10% glucose , pH 7 . 4 ) to a final concentration of 1 × 106 trypanosomes and injected intravenously into the sheep . Blood from ear vein was drawn to monitor parasite and PCV levels . All analyses were performed with fresh , undiluted wet blood films . Observation periods did usually not exceed 10 minutes . Videos of swimming trypanosomes were recorded at a frame rate of 500 fps using a Phantom camera v5 . 2 ( Vision Research , Wayne , NJ ) , mounted to either an automated iMIC microscope ( FEI Munich ) or a DM2500 microscope ( Leica microsystems ) , equipped with 60x and 100x objectives . Trajectories of trypanosome movement were traced using MTrackJ [26] . The trypanosomes were classified as described in the legend to Fig 1 . Speeds were calculated after measuring the translocation distance of persistent swimming phases . In order to simulate blood viscosity , 0 . 4% ( w/v ) methylcellulose ( Sigma-Aldrich ) was added to the cell culture [24] . A range of methylcellulose concentrations in TDB or mouse blood , from 0 . 2–0 . 8% w/v , was tested to investigate the effects of viscosity on trypanosome motion . Chemically inert polydimethyl siloxane ( PDMS ) -pillar arrays with defined diameters ( 8 , 10 and 12 μm ) and spacing ( 3 , 4 , 5 and 6 μm ) were used for analysis of trypanosome motion in confined geometries [24] . Trypanosomes were purified from infected mouse blood and applied in a volume of 10 μl of TDB ( trypanosome dilution buffer: 20 mM Na2HPO4 , 2 mM NaH2PO4 , pH 7 . 7 , 20 mM glucose , 5 mM KCl , 80 mM NaCl , 1 mM MgSO4 ) onto the pillar arrays . For single flagellar beat analysis ( Fig 6 ) , sequences were selected from high speed videos and processed with Fiji [54] . The oscillation of the flagellar tip was observed in successive frames ( 2ms intervals ) . At the beginning of each new full flagellar wave , the position of the posterior end of the cell body was measured ( white lines in Videos S8-S16 ) . The swimming speed was calculated after measuring the translocation distance in the direction of the cell´s movement and the beat frequency was calculated from the duration of each beat . For cellular waveform analysis ( Fig 4 ) , trypanosomes swimming persistently forwards with uninterrupted tip-to-base beats were selected and for the duration of one flagellar beat , the cell body was outlined in each frame in the 3d visualisation software Amira ( FEI ) . In Amira , the outlines were combined into a three-dimensional ( xyt- ) surface representation . Rotation of this surface model allowed the visualisation of the temporal dynamics of the cellular waveform as well as the oscillation pattern of the flagellar tip . For morphometric analyses live parasites were cell surface-labelled with 3 mM AMCA-sulfo-NHS ( sulfosuccinimidyl-7-amino-4-methylcoumarin-3-acetate , Thermo scientific , Pierce , Rockford ) essentially as described in [24] . The cells were fixed in a final concentration of 4% w/v formaldehyde and 0 . 25% v/v glutaraldehyde in 0 . 1 M HEPES buffer over night at 4°C . Fluorescent microscopy was done using the automated iMIC microscope equipped with a Pike camera ( PCO AG , Kelheim , Germany ) . The iMIC was controlled by Live Acquisition software ( FEI Munich , Germany ) . 3D models of fixed cells were computed from deconvolved high-resolution 3D image stacks image stacks ( z = 100 nm ) using the Huygens Essential Image processing software v4 . 3 ( SVI , Hilversum , Netherlands ) and the Amira software v5 . 6 . 0 ( FEI ) . An edge detection filter ( Sobel ) was applied and volume models were produced in Amira ( Voltex display ) . Cells were selected for surface modeling ( Isosurface display ) and completed with the function ′pointwrap′ . Flagella were traced using the volume model and Amira′s filament editor . The antibody clearance assay was essentially done as described [11] . Briefly , 108 trypanosomes were purified from murine blood by differential centrifugation ( 200xg , 5min , 4°C ) , resuspended in 0 . 5 ml of ice-cold Trypanosome Dilution Buffer ( TDB ) and cell surface-stained with 2 mM sulfo-NHS-SS-biotin ( Sigma-Aldrich ) for 15 min on ice . The parasites were washed twice in TDB and incubated on ice for 30 min with 10 μg/ml mouse monoclonal CF488A-conjugated anti-Biotin IgG antibody . Endocytosis was followed after warming 0 . 1 ml of cells to 37°C for specific time periods ( 20 s , 40 s , 1 min , 3 min , 5 min and 10 min ) . The process was stopped by chemical fixation in 0 . 8 ml of pre-warmed TDB , 2% ( w/v ) PFA . | African trypanosomes are protist flagellates that are successful parasites in a wide spectrum of hosts . These include humans , where they cause the deadly sleeping sickness , and livestock , where they cause nagana . Nagana has a tremendous negative impact in wide regions of sub-Saharan Africa . The motility of these parasites has been shown to be essential for their survival in all the different environments they inhabit , from the bloodstream of mammals to the gut of the tsetse fly vector . The complex swimming mechanism of trypanosomes has only recently been elucidated in detail , using Trypanosoma brucei cells that have been in long term culture . We aimed to characterise and compare the swimming behaviour of several important livestock-infective trypanosome species , isolated directly from the bloodstream . This was done using state of the art microscopy , allowing measurement of their motility with high spatiotemporal resolution . While showing that the basic flagellar propulsion mechanism is the same in all species , we related the trypanosomes motility to their characteristic morphology . We quantified distinct behaviours in the analysed species , which could specifically be manipulated by experimental variations in the physical environment . Importantly , we show that the trypanosome’s morphology and swimming performance could determine the anatomical niche the parasite populates in the host . This would allow differential dissemination of distinct trypanosome species in the mixed infections , which are frequently observed in the wild . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"motility",
"swimming",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"flagellar",
"motility",
"biological",
"locomotion",
"biomechanics",
"parasitic",
"protozoans",
"viscosity",
"protozoans",... | 2016 | Species-Specific Adaptations of Trypanosome Morphology and Motility to the Mammalian Host |
Recent evidence supports the involvement of inducible , highly diverse lectin-like recognition molecules in snail hemocyte-mediated responses to larval Schistosoma mansoni . Because host lectins likely are involved in initial parasite recognition , we sought to identify specific carbohydrate structures ( glycans ) shared between larval S . mansoni and its host Biomphalaria glabrata to address possible mechanisms of immune avoidance through mimicry of elements associated with the host immunoreactivity . A panel of monoclonal antibodies ( mABs ) to specific S . mansoni glycans was used to identify the distribution and abundance of shared glycan epitopes ( glycotopes ) on plasma glycoproteins from B . glabrata strains that differ in their susceptibilities to infection by S . mansoni . In addition , a major aim of this study was to determine if larval transformation products ( LTPs ) could bind to plasma proteins , and thereby alter the glycotopes exposed on plasma proteins in a snail strain-specific fashion . Plasma fractions ( <100 kDa/>100 kDa ) from susceptible ( NMRI ) and resistant ( BS-90 ) snail strains were subjected to SDS-PAGE and immunoblot analyses using mAB to LacdiNAc ( LDN ) , fucosylated LDN variants , Lewis X and trimannosyl core glycans . Results confirmed a high degree of glycan sharing , with NMRI plasma exhibiting a greater distribution/abundance of LDN , F-LDN and F-LDN-F than BS-90 plasma ( <100 kDa fraction ) . Pretreatment of blotted proteins with LTPs significantly altered the reactivity of specific mABs to shared glycotopes on blots , mainly through the binding of LTPs to plasma proteins resulting in either glycotope blocking or increased glycotope attachment to plasma . Many LTP-mediated changes in shared glycans were snail-strain specific , especially those in the <100 kDa fraction for NMRI plasma proteins , and for BS-90 , mainly those in the >100 kDa fraction . Our data suggest that differential binding of S . mansoni LTPs to plasma proteins of susceptible and resistant B . glabrata strains may significantly impact early anti-larval immune reactivity , and in turn , compatibility , in this parasite-host system .
Glycans are complex carbohydrate ( CHO ) chains normally covalently bound to polypeptides , lipids or other carrier molecules . Glycoconjugates such as glycoproteins , glycolipids and proteoglycans represent one of the most prominent classes of molecules exhibited by schistosomes . Schistosome glycans are highly diverse structurally and have been implicated in a variety of physiological processes during schistosome infection of its mammalian host , most notably their involvement in modulating protective immune responses and immunopathology ( see reviews [1]–[3] ) . Similarly glycans are also highly expressed in the free-swimming miracidial and intramolluscan developmental stages of Schistosoma spp . as shown by earlier exogenous lectin-binding studies [4] , [5] , and more recent glycotope/glycomic analyses [6]–[9] . However , despite the presence of diverse glycans associated with the larval surface and its secretions/excretions , their functional significance remains unknown . A popular notion that recently has gained traction in the Biomphalaria glabrata-Schistosoma mansoni system poses that larval glycans and/or their associated glycoconjugates may be serving as pathogen-associated molecular patterns ( PAMPs ) that interact with lectin-like pathogen recognition receptors ( PRRs ) , thereby mediating innate immune responses to invading miracidia ( see reviews [10]–[13] ) . This concept has been incorporated into a proposed mechanism , termed compatibility polymorphism [14] , in which it is hypothesized that high molecular diversity in relevant PAMP and PRR systems can provide the necessary variation in receptor-ligand interactions to account for differences in infection rates seen in different snail-schistosome strain combinations [15] . Two candidate gene families that fulfill the basic requirements of exhibiting high molecular polymorphism and potential functional diversity are the fibrinogen-related proteins or Freps , lectin-like proteins in plasma of B . glabrata snails [16] and a family of polymorphic mucins from S . mansoni ( SmPoMuc; [17] ) . Recent studies have reported the selective ability of SmPoMuc to form complexes with Freps from snail plasma [18] , as well as the demonstration of a direct linkage between expression of one B . glabrata Frep ( Frep 3 ) and resistance to trematode infection [19] , thus supporting a functional basis for the compatibility polymorphism hypothesis . The specific ligands mediating Frep-SmPoMuc binding , however , still remain unknown , although the lectin-like properties of Freps and the fact that SmPoMuc are highly glycosylated point to the likelihood that specific glycan structures may be serving to mediate larval recognition , leading to hemocytic encapsulation and parasite killing typical of resistant host phenotypes [20] , [21] . Taken further it may then be predicted that the absence of relevant snail Freps and/or differences in expressed glycans ( e . g . , based on variable SmPoMuc expression ) should lead to the opposite schistosome-snail outcome; that is , larval survival due to nonrecognition by the snail's immune system . Assuming snail-schistosome compatibility involves a diversified lectin-based immunorecognition system , the repertoire of larval-expressed glycans ( both qualitative and quantitative ) potentially has a direct impact on hemocyte reactivity towards a given parasite within the snail host . Following up on Damian's “molecular mimicry” hypothesis [22] , [23] , that a parasite constitutively expressing host-like molecules could render the host immunologically “blind” to the parasite's presence , it has been suggested that early invading larvae ( miracidia , sporocysts ) possessing shared antigenic structures with their host snail may evade recognition by the internal defense system [24] , [25] . Although earlier studies demonstrated serological reactivity between larval S . mansoni and snail host hemolymph [26] , [27] , Dissous et al . [28] were the first to show that shared CHOs are represented among those immunoreactive epitopes . Recent structural analyses of N-glycans from B . glabrata plasma ( cell-free hemolymph ) provide definitive evidence that glycan structures , specifically terminal fucosylated LacdiNAc variants and core-linked xylose are shared between S . mansoni and its snail host [7] . In follow-up studies using highly specific monoclonal antibodies ( mABs ) to these , and other CHO epitopes ( glycotopes ) , extensive crossreactivity has now been confirmed between larval glycans and those of various B . glabrata tissues [7] , [9] , [29] , notably between host hemolymph and proteins released during larval transformation [9] . During the first hours following miracidial entry into the snail host , a complex molecular interplay takes place in which an array of macromolecules are released during miracidium-to-sporocyst transformation [30] , [31] . As a consequence , newly developing primary sporocysts are enveloped in a glycan-rich localized environment comprised mainly of glycoproteins , but also may comprise other glycoconjugates . These larval transformation products or LTPs [31] , in addition to serving as a passive source of host mimicked” molecules , also may actively bind snail lectins ( e . g . , Freps; [16] ) , thereby blocking lectin reactivity against newly developing sporocysts [32] . Given the possible immune modulating effects of LTPs released at a critical time when schistosome miracidia/sporocysts are in the process of establishing infections in the snail host , the present study investigated the effect of LTP exposure on the profile of shared glycotopes associated with plasma from susceptible ( NMRI ) and resistant ( BS-90 ) strains of B . glabrata . Results of these experiments demonstrate that glycoconjugates released during S . mansoni larval transformation significantly alter patterns of shared plasma protein glycotopes by either binding and blocking , or by exposing them , thereby providing a possible mechanism by which molecules released by early developing larvae may impact initial immune interactions at the host-parasite interface .
All experimental protocols involving mice and rabbits used in the course of this study were reviewed and approved by the Institutional Animal Care and Use Committee ( IACUC ) at the University of Wisconsin-Madison under Animal Welfare Assurance No . A3368-01 . Hemolymph was obtained from S . mansoni-susceptible and -resistant B . glabrata snails ( NMRI and BS90 strains , respectively ) by the headfoot retraction method [33] . Hemolymph from approximately 150 snails ( 12–15 mm shell diameter ) of each strain were pooled and represented a common source for subsequent Western and far-Western blot analyses . During the course of this study a second replicate pool of hemolymph was obtained from which immunoblot analyses were repeated . Upon collection , hemolymph of each strain was dispensed into 1 . 5 mL microcentrifuge tubes containing cold , sterile Chernin's balanced salt solution ( CBSS; [34] ) creating a 1∶1 dilution of CBSS:hemolymph . Tubes were centrifuged at 260× g ( Eppendorf 5810R , Hauppauge , NY ) for 10 min to pellet hemocytes followed by transfer of the cell-free hemolymph ( plasma ) to a 15-mL Amicon centrifugal ultrafiltration tube ( Amicon Ultra-100 k; Millipore Corp , Billerica , MA ) with a nominal molecular weight cut-off of 100 kDa . Tubes were centrifuged at 1600× g for 95 min at 4°C ( Eppendorf 5810R ) . This plasma ultrafiltration step was performed in order to separate the snail hemoglobin ( which comprises the predominant protein constituent of hemolymph ) from a hemoglobin-depleted lower molecular weight plasma fraction . Fractions were designated “>100 kDa-fraction” , which contained the vast majority of hemoglobin as indicated by its intense red color , and “<100 kDa-fraction” , which typically was colorless . Concentration of the <100 kDa-fraction was then carried out using a 3 kDa MW cutoff centrifugal ultrafiltration tube ( Amicon Ultra-3 k; Millipore Corp , Billerica , MA ) . Protein concentrations were determined using a Nanodrop spectrophotometer ( ND-1000; Nanodrop Technologies , Wilmington , DE ) , followed by addition of protease inhibitors ( Protease Inhibitor Cocktail SetIII , EDTA-free; Calbiochem , San Diego , CA ) to protect proteins from endogenous protease activities . Miracidia were obtained from eggs recovered from infected mice provided by the Biomedical Research Institute ( BRI , Rockville , MD ) and housed at the University of Wisconsin Charmany Instructional Facility . Miracidia were isolated under axenic conditions as previously described [35] and placed into in vitro culture in CBSS supplemented with penicillin and streptomycin and 1 g/L each of glucose and trehalose ( CBSS+ ) under normoxic conditions at 26°C . Culture supernatants containing larval transformation products ( LTPs ) were harvested after 24 hr by filtering through a Whatman Puradisc 0 . 2-µm syringe filter ( GE Healthcare LTD , Buckinghamshire , UK ) to remove stray sporocysts , epidermal plates , and any cellular debris . Culture supernatants were then concentrated by ultrafiltration ( Amicon Ultra-3 k; Millipore Corp , Billerica , MA ) and protein concentration determined using a Nanodrop ND-1000 . Following addition of protease inhibitors as described above , aliquots were stored at 4°C if used within a week , or at −80°C for longer storage periods . Before use in far-Western blot analyses , concentrated LTP was diluted in snail Tris-buffered saline ( sTBS; 20 mM Tris-HCl , 150 mM NaCl , pH7 . 4 ) .
Subjecting B . glabrata plasma to centrifugal molecular filtration using 100 kDa MW spin-filters generally was effective in separating intact plasma into fractions enriched for proteins above ∼75 kDa ( >100 kDa cut-off ) and those of 50 kDa and lower ( Figure 1A ) . Because these ultrafiltration membranes only provide a nominal MW cut-off , clearly there is considerable MW overlap between the high and low MW fractions . The overlap of mid-range MW proteins in both fractions ( 25–75 kDa ) also may have been the result of breakdown of high MW complexes in the >100 kDa fraction during SDS-PAGE processing . Importantly , however , based on protein banding patterns revealed in Coomassie blue- and silver-stained SDS-PAGE gels , only minor differences in protein molecular mass profiles were evident between NMRI and BS-90 snail strains . When separated and blotted snail plasma was probed with a polyclonal antiserum generated against products released during in vitro miracidial transformation ( LTP ) , there was extensive immunoreactivity with plasma in both high ( >100 kDa ) and low ( <100 kDa ) molecular weight plasma fractions ( Figure 1B ) . Reactivity of our anti-LTP polyclonal antiserum to blotted plasma provided a confirmation that plasma proteins did in fact share crossreactive antigens ( epitopes ) with products released by S . mansoni larvae during early larval development . Because few differences in crossreactive bands were observed between NMRI and BS-90 plasmas using the polyclonal antibody to undefined antigens contained in the LTP , we sought to incorporate highly specific mAB to defined epitopes to identify possible snail strain differences in shared determinants . The similarity in crossreactive plasma profiles also served as an additional sample loading control , allowing for comparisons of mAB reactivities between snail strain samples . These initial analyses of snail plasmas established that ( 1 ) NMRI and BS-90 plasma have similar protein content ( loading control ) , ( 2 ) that plasma and LTPs did , in fact , share an extensive array of undefined epitopes , and ( 3 ) that the use of specific mABs to defined epitopes could provide for a detailed comparison of the shared epitopes exhibited by NMRI and BS-90 snail plasmas in the presence and absence of LTP . In contrast to polyclonal anti-LTP staining , the immunoreactivities of 8 glycan-specific mABs with native ( untreated ) B . glabrata plasma proteins were highly varied as shown by Western blot analyses . Highest reactivity , both in staining intensity and number of proteins , was observed using the trimannosyl N-glycan core ( TriMan ) mAB , while no staining was seen with anti-Lewis X ( LeX ) ( Figure 2; −LTP ) . Snail strain differences in staining patterns for TriMan were detected including greater reactivity with NMRI proteins between 50–75 kDa ( Figure 2; −LTP/<100 kDa fraction ) and in BS-90 , proteins between 15–20 kDa ( Figure 2; −LTP/<100 kDa fraction ) . Moreover , pre-incubation of the blotted plasma proteins with S . mansoni LTP had little effect on the patterns of shared TriMan glycotopes , with the exception of enhanced staining of BS-90 proteins in the 25–37 kDa range for TriMan ( Figure 2; +LTP/>100 kDa fraction ) . Similarly , interactions of LTP with high MW BS-90 plasma proteins resulted in the appearance of anti-LeX-reactive bands between 50–150 kDa ( Figure 2; +LTP/>100 kDa fraction ) . Similar to TriMan mAB reactivity , consistent qualitative and/or quantitative differences in shared glycotope distribution between native ( untreated ) NMRI and BS-90 plasma proteins were observed for LDN , F-LDN and F-LDN-F glycotopes , especially in the <100 kDa fraction ( Figure 3; −LTP/<100 kDa fraction ) . Typically , NMRI exhibited higher levels and/or wider distribution of these glycotopes when compared to BS-90 proteins , and notably displayed unique patterns of glycotope reactivity on native proteins indicating that shared glycotopes on NMRI proteins were either associated with unique subsets of glycoproteins or differed quantitatively in their occurrence on individual plasma proteins . Although the occurrence of LDN , F-LDN , LDN-F and F-LDN-F glycotopes varied considerably on plasma proteins contained in the >100 kDa fraction ( Figure 3; −LTP/>100 kDa ) , in the absence of LTP treatment , no consistent snail strain differences were observed for these higher MW proteins . Monoclonal antibodies to LDN-DF and DF-LDN-DF exhibited little reactivity against native plasma proteins , with the exception of higher MW polypeptides in the >100 kDa fraction that stained with anti-DF-LDN-DF ( Figure 4; −LTP/>100 kDa ) . Also , no notable snail strain differences in the distribution of these glycotopes were observed . Exposure of SDS-PAGE-separated and blotted snail plasma proteins to LTP had a profound effect on the reactivity of plasma proteins to the panel of anti-glycan mABs . Alterations in the exposed glycotope distribution on plasma proteins were manifested in several ways: i ) decreased glycotope-specific antibody reactivity in selected plasma proteins as noted with F-LDN , F-LDN-F and DF-LDN-DF ( Figures 3 and 4; +LTP/>100 kDa fraction ) , ii ) qualitative changes in antibody reactivity to plasma proteins when compared to non-LTP exposed blots , e . g . , F-LDN and F-LDN-F ( Figure 3; +LTP/<100 kDa fraction ) or iii ) appearance of specific glycotopes in previously undetected or existing protein bands . These include the reactivities of anti-LDN-F , LDN-DF and DF-LDN-DF in LTP-exposed plasma proteins in the <100 kDa fraction ( Figs . 3 and 4; +LTP/<100 kDa ) and anti-LeX , LDN , LDN-F and LDN-DF mAB reactions to the >100 kDa fraction ( Figures 2 , 3 and 4; +LTP/>100 kDa ) . Furthermore , it was noted that exposure of blotted NMRI and BS-90 plasma to LTP altered the native ( non-treated ) glycotope staining profiles for LDN-F , F-LDN , F-LDN-F , and LDN-DF ( Figures 3 and 4; +LTP/<100 kDa ) and LeX ( Figure 2; +LTP/>100 kDa ) , in a snail strain-specific manner . Examples of plasma proteins that exhibit snail strain-specific changes in glycotope immunoreactivity following LTP-exposure are listed in Table 2 .
As alluded to in the introduction , the importance of glycans as potential PAMPs in this model , and probably other trematode-snail systems , has been highlighted in recent studies focusing on the identification of specific glycan structures associated with early developing larval stages of S . mansoni [6] , [8] , [9] , [29] . In the present study , mABs generated from Schistosoma-infected mice and reactive for terminal elements of schistosome glycans expressed by miracidial and primary sporocyst stages [6] , [8] , [9] , exhibited extensive reactivity with cell-free hemolymph ( plasma ) glycoproteins of B . glabrata . In addition , based on the plasma protein binding patterns for selected glycotope-specific mABs , clear differences were observed between the NMRI and BS-90 snail strains; most notably NMRI plasma possessed a greater number and/or higher staining intensity of glycoproteins displaying the LDN , F-LDN and F-LDN-F glycotopes than plasma proteins from BS-90 snails . These findings are generally consistent with those reported by Lehr et al . [29] , although in the present study no consistent snail-strain differences were observed for shared LDN-F , LDN-DF , DF-LDN-DF on native plasma proteins . Based on the results of these shared glycotope studies , a main goal of the present work was to determine if molecules released during early S . mansoni larval development ( miracidium-to-sporocyst transformation ) are capable of interacting with plasma proteins , and if so , what effect this interaction may have on the patterns of shared plasma glycans in these B . glabrata strains . The initial 24 hours of schistosome larval development is believed to be a critical period in determining the success or failure of establishing infections within the snail host [12] . Therefore , unraveling the complex molecular interactions occurring during this time should provide important insights into the underlying mechanisms of host-parasite compatibility in this model system . Upon penetration of a susceptible snail host , miracidia rapidly ( within hours ) begin transforming to sporocysts by shedding their ciliated epidermal plates during the process of sporocyst tegument formation [40] , [41] . As larvae transform they release a complex mixture of molecules ( LTPs ) , including many glycoproteins [30] , [31] , [42] , which are capable of binding snail hemolymph components [43] , [44] and modulating hemocyte function [13] , [45]–[47] , likely through the action of mimicked CHOs [48] . Since innate resistance to incompatible S . mansoni strains is mediated by hemocytic encapsulation , the molecular composition of LTPs confronting immune elements in hemolymph would be predicted to have functional consequences resulting from LTP binding to plasma or hemocyte proteins , thereby interfering with larval recognition mechanisms , or by altering hemolymph glycoprotein structures through enzymatic means . In the present study , we have shown that exposure of plasma proteins to LTP dramatically alters the patterns of anti-glycan mAB reactivity for selected glycotopes naturally-occurring ( shared ) on snail plasma proteins . The most notable effects of LTP exposure on shared glycotope immunoreactivity include reduced or enhanced immunostaining of existing reactive protein bands ( e . g . , LDN , F-LDN ) , and changes in existing patterns of plasma glycotope reactivity that results in the appearance or disappearance of immunoreactive bands as noted for anti-F-LDN-F and LDN-F . At present the mechanism ( s ) by which LTPs alter plasma protein glycotope reactivity is not precisely known . However , two of the most likely possibilities include ( 1 ) the binding of LTP directly to subsets of blotted plasma proteins that results in either enhanced glycotope display if the bound LTP are glycoproteins , or decreased glycotope reactivity due to LTP binding and blocking mAB-reactive plasma glycans , or ( 2 ) the presence of specific glycosidases in LTP preparations that are capable of enzymatically altering glycan structures , thereby potentially changing the pattern of specific glycotope reactivity displayed by plasma proteins . A proteomic analysis of S . mansoni LTP recently identified an α-N-acetylgalactosaminidase and an endo-α-mannosidase [31] , both of which may have potential for altering or removing non-substituted GalNAc ( as part of LDN , LDN-F and possibly LDN-DF ) and TriMan structures , destroying their mAB-reactivities in the process . However , although specific glycosidase activities may account for LTP-associated decreases in plasma glycotope reactivity , their presence/activities cannot explain why exposure to LTP often results in enhanced mAB reactivity for selected plasma proteins ( e . g . , LDN , LDN-F ) or variable reactivity for a specific mAB among different proteins in the same plasma sample ( e . g . , F-LDN-F ) . Moreover , because the endo-α-mannosidase ( pH optimum , 7 . 0 ) only cleaves α1–2 linkages of extended oligomannose structures [49] , it is highly unlikely to be active against the α1–3 and α1–6 linkages comprising the TriMan N-glycan core . The S . mansoni α-N-acetylgalactosaminidase also is not predicted to cleave the β1–4 linkage of LDN , thereby rendering it incapable of altering LDN-associated glycotopes . Finally the pH optimum for this enzyme is between 4 . 3–4 . 8 [50] and likely would not be active at the pH of the LTP incubation medium ( pH 7 . 4 ) . Therefore , based on the current data , we hypothesize that the direct binding of specific LTPs to selected plasma proteins is responsible for the observed changes in plasma glycotope staining patterns . At this point , however , we do not know to what extent this LTP-plasma binding is due to protein-CHO or protein-protein interactions . Because LTPs appear to bind selectively to subsets of plasma proteins , and in many cases resulting in significant increases or decreases in shared glycotopes associated with a LTP-protein complex , one can imagine a scenario in which miracidia undergoing transformation within the snail host release LTPs that selectively bind to plasma proteins that may be serving as PRRs . Indeed , many of the changes in plasma immunoreactivity seen in LTP-exposed blots occur in proteins between 50–100 kDa ( e . g . , LDN , LDN-F , F-LDN-F , LDN-DF ) , which corresponds generally to the molecular size range of the major B . glabrata Freps [16] . Current efforts are being focused on identification of the host hemolymph proteins involved in LTP binding interactions . The findings reported here , and in other recent investigations ( reviewed in [12]–[14] ) present a number of unanswered questions regarding the functional significance of glycans shared between and/or manipulated within schistosome-snail systems . According to the compatibility polymorphism hypothesis detailed by Mitta et al . [14] , the maintenance of high molecular polymorphism ( genetic diversity ) in both host PRRs ( e . g . , lectin-like proteins ) and their corresponding parasite PAMPs ( e . g . , glycan-bearing polymorphic mucins ) may serve as a possible co-evolutionary mechanism driving the compatibility/incompatibility phenotypes observed in this , and other snail-trematode systems . Although the mechanisms by which snail hemocytes first recognize invading schistosome larvae ( including both cellular attraction and adhesion/encapsulation ) are still unknown , members of a highly diversified family of fibrinogen-related proteins ( Freps ) have been implicated as anti-trematode PRRs [19] , [51] , [52] . Since it has been shown that lectin-like Freps utilize sugars as recognition ligands [19] , glycan structures naturally shared between larval S . mansoni and host hemolymph or parasite molecules that alter the composition of native glycans displayed by plasma glycoproteins may serve an important role in determining the reactivity of hemocytes towards an individual parasite developing within a given host [12] , [13] , [24] . With reference to the parasite , genes encoding the polymorphic mucin family ( SmPoMuc; [17] , [32] ) are proposed to provide sufficient genetic diversity to maintain an adequate counterbalance to the highly diversified snail Freps [14] . This is a reasonable notion , especially when one takes into account the added ligand diversity provided by the O- and N-linked glycan structures that can vary greatly , both qualitatively and quantitatively [9] , [29] . This glycan diversity was well illustrated in our study showing that , not only are parasite glycans ( glycotopes ) naturally found on numerous plasma proteins , but snails differing in their susceptibility to the S . mansoni strain used in this study , also differed in their repertoires of plasma-associated glycans ( e . g . , LDN and F-LDN ) . Moreover , snail-strain differences in the distribution of shared glycans also were noted following LTP treatment of plasma blots ( e . g . , LeX , F-LDN , LDN-F , LDN-DF ) . What is emerging from the present , and other related , investigations is a complex picture of host-parasite molecular interactions that , in accordance with the compatibility polymorphism hypothesis , should provide important clues as to which combinations of host immune factors and corresponding larval factors determine infection success or failure . In support of the above hypothesis , Mone et al . [18] found that fibrinogen-related protein Frep2 bound SmPoMuc in a coimmunoprecipitated complex , which also included , among other adhesion molecules , a Gal-binding lectin . Identifying the repertoire of host immune PRRs ( specifically in cell-free plasma and circulating hemocytes ) and their specific target ligands ( tegumental surface- and LTP-associated glycans ) represents the next critical step towards gaining a more comprehensive understanding of the molecular basis for compatibility and incompatibility in this laboratory model of molluscan schistosomiasis . As in mammalian host immune responses to schistosomes [2] , [3] , it is predicted that glycans , especially those shared with the host , will be shown to play prominent roles as modulators and/or mediators of larval-snail immune interactions . | Early larval stages of the human blood fluke Schistosoma mansoni face many barriers in their attempt to establish successful infections within their snail host , Biomphalaria spp . The snail's internal defense system represents one such barrier , which includes lectin-like recognition receptors and circulating hemocytes capable of encapsulating and killing invading larvae . Since host lectins likely are involved in early immunorecognition events , the recent identification of specific carbohydrate structures ( glycans ) shared between larval stages of S . mansoni and its host Biomphalaria glabrata suggests that larvae may be avoiding immune recognition through a molecular mimicry mechanism mediated by lectin-reactive glycans . Results of the present study support previous findings of extensive host-parasite glycan sharing , and demonstrate that molecules released by S . mansoni miracidia during in vitro development ( larval transformation products or LTPs ) selectively bind to plasma proteins , altering their reactivity to various glycan-specific monoclonal antibodies . Moreover , some of the changes in recognized glycans following exposure of blotted plasma proteins to LTP were B . glabrata strain-specific . We hypothesize that the differential interaction of LTPs with plasma proteins from different B . glabrata strains may play an important role in influencing the efficacy of anti-larval immune reactivity within a given host strain . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"immunology",
"biology",
"molecular",
"cell",
"biology"
] | 2012 | Glycotope Sharing between Snail Hemolymph and Larval Schistosomes: Larval Transformation Products Alter Shared Glycan Patterns of Plasma Proteins |
The role of Activation-Induced Cytidine Deaminase ( AID ) in somatic hypermutation and polyclonal antibody affinity maturation has not been shown for polyclonal responses in humans . We investigated whether AID induction in human B cells following H1N1pdm09 vaccination correlated with in-vivo antibody affinity maturation against hemagglutinin domains in plasma of young and elderly individuals . AID was measured by qPCR in B cells from individuals of different ages immunized with the H1N1pdm09 influenza vaccine . Polyclonal antibody affinity in human plasma for the HA1 and HA2 domains of the H1N1pdm09 hemagglutinin was measured by antibody-antigen complex dissociation rates using real time kinetics in Surface Plasmon Resonance . Results show an age-related decrease in AID induction in B cells following H1N1pdm09 vaccination . Levels of AID mRNA before vaccination and fold-increase of AID mRNA expression after H1N1pdm09 vaccination directly correlated with increase in polyclonal antibody affinity to the HA1 globular domain ( but not to the conserved HA2 stalk ) . In the younger population , significant affinity maturation to the HA1 globular domain was observed , which associated with initial levels of AID and fold-increase in AID after vaccination . In some older individuals ( >65 yr ) , higher affinity to the HA1 domain was observed before vaccination and H1N1pdm09 vaccination resulted in minimal change in antibody affinity , which correlated with low AID induction in this age group . These findings demonstrate for the first time a strong correlation between AID induction and in-vivo antibody affinity maturation in humans . The ability to generate high affinity antibodies could have significant impact on the elucidation of age-specific antibody responses following vaccination and eventual clinical efficacy and disease outcome .
Antibody affinity maturation is a key aspect of an effective immune response to vaccines likely to provide a significant protection against human pathogens . The discovery of Activation-Induced Cytidine Deaminase ( AID ) has led to the elucidation of key molecular mechanisms involved in class switch recombination ( CSR ) and somatic hypermutation ( SHM ) , which occur in B cells as they mature in germinal centers of lymph nodes and spleen in response to antigenic stimulation and T cell signals [1]–[2] . Recently , using mouse models of AID-dependent cell labeling , it was shown that memory B cells appear in the IgM+ and IgG+ subsets both in germinal centers and outside of B cell follicles . After challenge , the IgG+ memory B cells differentiate into plasma cells , whereas the IgM+ memory B cells reinitiate a germinal center reaction , resulting in class switching and SHM leading to production of higher affinity BCR expressed on memory and plasma cells [3] . Age-related defects in B cells have been reported and these include decrease in AID expression due to impairment of the transcription factor E47 which activates AID [4] . Recirculating B cells found in the peripheral blood can be used to measure response under in vitro conditions that may partially reflect the in vivo vaccine-induced immune response of the individual . Following polyclonal or antigen-specific stimulation in vitro , AID expression was shown to increase in B cells from young individuals , but significantly less in B cells from elderly individuals . This up-regulation of AID expression occurs primarily in stimulated naïve and IgM+ memory , but not in switched memory B cells [4] , [5] , [6] . More recently , in two studies on age effects on influenza vaccination , we have shown that levels of AID in vaccine-stimulated B cells and serum antibody responses are positively correlated in humans [6]–[7] . This correlation suggested that a defect in AID induction in PBMC derived B cells ( measured in vitro ) correlated with the observed decrease in the number of switched memory B cells and a reduced number of IgG plasmablasts after vaccination in vivo [6]–[7] . While association between CSR and AID has been suggested before , no information to date is available for the direct involvement of AID and SHM and/or antibody affinity maturation in humans . Influenza subtypes are classified based on the antigenic variation within influenza hemagglutinin ( HA ) as measured by a hemagglutination inhibition ( HI ) assay . The HI assay is dependent on antibodies that inhibit the interaction between the sialic acid receptor on red blood cells ( RBC ) and the receptor binding domain ( RBD ) within the globular domain ( HA1 ) of influenza hemagglutinin . Therefore , the antigenic differences within influenza viruses are primarily due to mutations within the HA1 domain , while the protein sequence within the HA2 stalk domain is highly conserved among multiple influenza subtypes and strains . Human polyclonal responses against one subtype can show significant cross-reactivity to the hemagglutinin of other subtypes due to this high sequence conservation in the HA2 domain as previously shown [8] , [9] , [10] . But this binding cross-reactivity does not commonly translate into cross protection , since most of the antibodies against the HA2 stalk do not block virus infectivity . In our previous studies , we have demonstrated that most of the polyclonal neutralizing antibody responses following influenza infections or inactivated subunit vaccination , as measured in HI or microneutralization ( MN ) assays , targeted the HA1 domain [8] , [11] . Recently , rare antibodies with broad neutralizing cross-reactivity that target the HA2 stem were reported , but they are not easily elicited by traditional vaccination [12] . Therefore it is important to study the humoral responses against the different domains within the influenza hemagglutinin that were shown to evolve independently for the HA2 and HA1 antigenic regions [10] . The involvement of AID in CSR has been shown . However , the involvement of AID in somatic hyper mutation ( SHM ) , leading to antibody affinity maturation , is poorly understood especially for polyclonal antibody responses in humans . In the current study , we correlated for the first time the levels of AID in peripheral B cells and AID fold increase post-H1N1pdm09 vaccination with the in vivo humoral immune responses against the H1N1pdm09 inactivated influenza vaccine in individuals ranging in age from 20 to 90 years . We measured polyclonal antibody affinity of human plasma to the globular domain ( HA1 ) and the conserved stalk domain ( HA2 ) of the H1N1pdm09 hemagglutinin using a Surface Plasmon Resonance ( SPR ) based real-time kinetics assay as previously described [9] , [10] . Changes in polyclonal antibody-antigen complex dissociation rates , as indicators of antibody affinity maturation in human plasma following vaccination were correlated with fold-increases in AID levels of B cells after vaccination and with AID levels in B cells before vaccination . These results demonstrate for the first time that AID contributes to polyclonal antibody affinity maturation in response to influenza vaccination in humans .
Forty two individuals , 20–90 year old , were enrolled in the H1N1pdm09 vaccine trial . HI titers and AID mRNA expression were evaluated at day 0 ( t0 ) and day 28 ( t28 ) ( before vaccination and 28 days after vaccination ) . All the individual data are presented in Table S1 . Only 3/26 young and 2/16 elderly individuals had non-protective titers at t0 ( below 1∶40 ) . Moreover , 2/26 young and 3/16 elderly had higher HI titers at t0 ( 1∶160–1∶640 ) , reflecting possible recent exposure to the first wave of the pandemic H1N1pdm09 . Importantly , vaccination induced a significant boost in HI titers that was more robust in young compared with elderly individuals ( average fold-increase of 13 in young versus 4 in elderly individuals , respectively , p = 0 . 004 ) ( Table S1 ) . Although it has been reported that archived plasma samples from elderly ≥80 years of age have increased neutralization titers to the H1N1pdm09 [13] , the initial titers in young and elderly individuals in our cohorts were not significantly different ( reciprocal titers at t0: 74±12 and 108±39 in young versus elderly , respectively , p = 0 . 44 ) . Therefore , the response parameters measured , including AID fold-increase at t28 , are not likely to be impacted by difference in initial ( t0 ) HI titers , which reflect long lived plasma cells . We have previously shown that the intrinsic defect in AID expression in B cells from the elderly population is a general phenomenon , and therefore not limited to influenza specific cells [14] [4] . In the current study , we have used the in vitro AID response as a biomarker for an effective influenza-specific B cell response . The inactivated vaccine used for vaccination of the subjects was also used for in vitro stimulation at t0 and t28 . This approach is supported by our earlier studies demonstrating a strong correlation between AID inducibility in peripheral B cells in vitro with the in vivo serum HI responses [6]–[7] . A two fold change ( or higher ) in AID mRNA level was empirically determined as a positive response to the vaccine and was found to correlate with positive HI responses [4]–[6] . Results shown in Table S1 and Fig . 1 demonstrate lower fold-increase in AID mRNA levels in peripheral B cells following H1N1pdm09 vaccination in elderly individuals ( 3/16; 19% , with AID fold change ≥2 ) compared with the younger adults with 22/26 individuals ( 85% showing ≥2 fold increase in AID mRNA expression at t28 ) . A statistically significant negative correlation coefficient was observed between levels of AID induction and age ( Fig . 1 ) . Furthermore , a positive correlation was observed between fold increase in AID induction ( t28/t0 ) and the fold increase in HI titers ( Table S1 ) , confirming our previous results showing that the in vitro AID mRNA increase in vaccine-stimulated B cells positively correlated with the influenza vaccine-specific HI titers [6] , [7] . These results are not due to higher numbers ( and their phenotypes ) or the percentages of memory B cells ( both switched memory and IgM memory ) in elderly individuals , as we have previously demonstrated that these do not increase with age . Since AID induction occurs primarily in naïve and IgM memory cells , the findings in the current cohort confirm our previous observations on the age-related decline in AID induction in peripheral B cells , and are not simply a reflection of a decrease in the number of responding B cells . Also , this cannot be due to a difference in kinetics in B cell responses , because we have done kinetics previously in young and elderly and found the peak response to be coincident and optimal at t28 for both age groups [6] . To determine the effect of age on antibody affinity maturation in human plasma against H1N1pdm09-HA following vaccination , polyclonal antibody off-rate constants , which describe the strength of antigen-antibody complexes , were determined directly from the antibody interactions with the HA1 globular domain or HA2 stalk domain , using SPR as described before [9] . We have previously established that the antigen-antibody dissociation kinetics is not influenced by the antibody concentration in the polyclonal sera but reflects the avidity of bound antibodies for the proteins on the chip surface . For each individual , the dissociation off-rates ( Kd ) were measured for pre-vaccination day 0 ( t0 ) and post-vaccination day 28 ( t28 ) samples ( Fig . 2A ) , and fold-change in Kd to the HA1 ( Fig . 2B–C ) and HA2 domains ( Fig . 2D ) were calculated . Positive fold change of antibody off-rates ( post/pre-vaccination values ) indicates slower dissociation of antigen-antibody complexes reflecting increased polyclonal antibody affinity in the plasma of vaccinated individuals . Since the HA1 globular domain is not as conserved as HA2 , the response to HA1 better estimates the newly generated strain-specific antibodies , and is better suited for measurements of affinity maturation following H1N1pdm09 vaccination . In most subjects , following vaccination , the anti-HA1 antibody dissociation rates decreased ( indicating increased antibody affinity ) by ≥1 log for most vaccinated individuals , reaching off-rates of 10−3–10−4 per sec ( Fig . 2A; filled vs . open circles ) . Interestingly , in a significant fraction of older adults ( ≥65 years ) , the pre-vaccination ( t0 ) anti-HA1 antibody binding off-rates were significantly slower ( <10−3/sec ) in the plasma compared with the rest of the cohort ( >10−3/sec ) , demonstrating the presence of high affinity antibodies against the H1N1pdm09 HA1 prior to vaccination ( Fig . 2A , open circles ) in that age group . Our data also demonstrated an inverse correlation between the fold-change in anti-HA1 antibody affinity following H1N1pdm09 vaccination and the pre-vaccination antibody off-rates to the HA1 globular domain ( Fig . 2B ) . Importantly , the fold increase in polyclonal antibody affinity following H1N1pdm09 vaccination ( t28/t0 ) to the HA1 domain , as measured by slower off-rates , declined with age ( Fig . 2C ) . In contrast to the findings with HA1 coated chips , the affinity of polyclonal antibody binding to the HA2 stalk domain , which is highly conserved between the H1N1pdm09 and seasonal H1N1 influenza strains , was uniformly high ( off-rate constant ∼10−4/sec ) even before vaccination , and the fold change in anti-HA2 antibody off-rates ( t28/t0 ) of human plasma did not demonstrate any age dependence ( Fig . 2D ) . To determine if AID induction following H1N1pdm09 vaccination correlates with the change in antibody affinity against HA1 and HA2 domains of H1N1pdm09-HA , the increase in AID expression in stimulated B cells following H1N1pdm09 vaccination ( t28/t0 ) was correlated with the fold change in antigen-antibody complex off-rate binding constants to the HA1 globular domain ( Fig . 3A ) or HA2 stalk domain ( Fig . 3B ) , for each vaccinated individual . As can be seen in Fig . 3 , a strong direct correlation was demonstrated between the fold increase in AID and the increase ( fold change ) of antibody affinity to the HA1 domain ( r = 0 . 867 ) ( Fig . 3A ) . A significant positive correlation was also observed between AID levels at t0 and fold increase in antibody affinity to the HA1 domain ( Fig . 3C ) , indicating that both AID levels in B cells at t0 and the capacity of B cells to up-regulate AID expression during an antigen-specific response are likely contributors to antibody affinity maturation in-vivo . In contrast , no correlation between the change in antibody binding affinity to the conserved HA2 stalk domain and AID activity was observed ( Fig . 3B ) .
The involvement of AID in class switch recombination ( CSR ) has been shown [15] , [16] . However , the role of AID in somatic hypermutation and antibody affinity maturation has only been postulated in humans , and may be extremely important in understanding age-related antibody responses following vaccination and their impact on vaccine efficacy and disease outcome . Germinal centers ( GC ) in lymph nodes and spleen were shown to be the key anatomical sites where B cells undergo class-switch and affinity maturation in AID-dependent mechanisms after exposure to foreign antigen . The GC formation depends on close interactions between antigen specific B and follicular helper T cells ( TFH ) [17]–[18] [19] , [20] . In humans it is very difficult to access the lymph nodes for large scale studies . However , there is evidence that circulating peripheral B cells could be used to measure AID mRNA baseline levels and induction after vaccination , thus mimicking the earlier GC reactions . We have previously shown that AID can be induced in cultures of B cells stimulated not only with anti-CD40/CD40L plus cytokines [14] [4] , but also with vaccines and CpG [6]–[7] . Thus , the response of peripheral B cells to vaccine stimulation in vitro in the presence of co-stimulatory signals provided by T cell signals or TLR agonist seems to mimic the early germinal center reactions . Using AID mRNA induction as a parameter of vaccine responsiveness , we have previously shown that the in vivo and in vitro response of B cells to the influenza vaccine are both decreased with advanced age with strong correlation , providing support for the use of AID induction in peripheral B cells as a biomarker of in vivo immune responsiveness [6]–[7] . In previous studies we have compared the response of purified B cells with PBMC and the AID fold-increase to the vaccine is comparable . Therefore , the contribution of T cells to this AID response is not significant [7] . In the current study we demonstrate for the first time a strong correlation between baseline AID levels as well as fold-induction of AID activity following vaccination and polyclonal antibody affinity maturation measured by SPR in plasma following influenza vaccination in humans . We also demonstrate the presence of pre-existing high affinity antibodies to the pandemic H1N1 HA1 globular domain in a fraction of older adults ( 65–90 yr ) in agreement with our findings in another H1N1pdm09 vaccine trial [10] . On the other hand , the affinity of polyclonal antibody binding to the H1N1pdm09 HA2 stalk domain was high even before vaccination ( ∼10−4/sec ) in all individuals , irrespective of age , and did not increase post-vaccination . The high conservation of HA2 protein sequence between H1N1pdm09 and circulating seasonal H1N1 influenza strains most likely have resulted in maximal affinity maturation of HA2-binding antibodies . Therefore , even in younger individuals with significant induction of AID expression post vaccination , there was no significant increase in the affinity of anti-HA2 antibodies in human plasma . These findings suggest that there is an upper limit of antibody affinity for human polyclonal antibodies in-vivo . In a recent study on B cell responses to seasonal Trivalent Inactivated Vaccine ( TIV ) , it was found that the frequency of vaccine-specific plasmablasts were lower in the elderly than in young adults . However , no clear difference in vaccine-specific affinity of plasmablast derived polyclonal or monoclonal antibodies against complete hemagglutinin ( HA0 ) was found between younger and older individuals [21]–[22] . These findings are of particular importance , since in these studies the entire HA vaccine was used to measure total antibody secreting cells ( ASC ) and binding avidity of antibodies . Due to the high conservation of HA2 , the level of cross-reactive binding to the stalk domain could have masked significant differences in the avidity of binding to the HA1 domain among different age groups . Moreover , AID activity was not measured in the previous study and the role of AID in promoting antibody affinity maturation in polyclonal B cell responses in humans following vaccination was not understood . In other studies , the frequency of mutations in Ig VH/VL genes in peripheral and tonsillar B cells of older people was increased when compared to younger individuals[23]–[24] , but this did not correlate with AID expression levels , suggesting that the hypermutated Ig VH/VL genes were from long lived ( affinity matured ) memory B cells and plasma cells . The affinity of antibodies is likely to play a key role in vivo , especially very early in infection ( i . e . , unfavorable antibody/viral load ratios ) . During the H1N1pdm09 pandemic , low affinity antibodies in some infected individuals were associated with more severe disease [25] . The high affinity anti-HA1 antibodies found in some older individuals ( ≥65 years ) prior to vaccination most likely represent long term plasma cells or memory IgG+ B cells that have undergone affinity maturation through exposure to H1N1pdm09 like influenza virus strains that circulated in the US during the first half of the 20th century . Importantly , the significant differences in binding affinity found between the older vs . younger adults was only demonstrated for antibodies targeting the HA1 globular domain but not the HA2 stalk domain . The presence of higher affinity antibodies to HA1 in older individuals pre-vaccination provides an additional explanation for the unusually low rate of severe respiratory disease during the 2009 H1N1 pandemic in this age group that is usually most susceptible to morbidity and mortality due to seasonal influenza . Thus , the present study demonstrated that an increase in AID expression was associated with affinity maturation of polyclonal antibodies in human plasma . In the older individuals , AID induction was very low . However , in spite of low AID activity and minimal change in and HI titers , post-H1N1pdm09 vaccination , older adults were relatively protected from the pandemic H1N1 compared with younger adults probably due to the presence of affinity mature B cells and polyclonal antibody responses [26] , [27] , [28] , [29] , [30] . We hypothesize that the antibody responses to the influenza vaccine during the 2009 and 2010 seasons in elderly ( in terms of antibody affinity ) will not be the norm when novel influenza strains are circulating , which is the more common scenario . The ability to generate new high affinity antibodies from naïve B cells will require AID activation following exposure to neoantigens such as drifted seasonal influenza strains , avian influenza ( H5N1 , H7N7 ) , or emerging pathogens , putting the older individuals at a disadvantage [31] , [32] . Therefore , vaccination of divergent human populations , especially older individuals , should take into consideration the AID status and the history of exposure and vaccination against the specific pathogen . Targeted vaccination of older adults ( e . g . , against varicella reactivation and other pathogens ) is likely to be more effective before the precipitous drop in AID activity .
Experiments were conducted using blood isolated from healthy volunteers of different ages after appropriate signed informed consent . The study has been approved with IRB protocol #20070481 . Participants included 26 young subjects ( age 20–59 years ) and 16 elderly ( age 60–90 years ) . Of these , 22 young and 12 elderly were vaccinated using the H1N1pdm09 monovalent vaccine ( from Novartis ) in the 2009–10 season , whereas 4 young and 4 elderly were vaccinated with the trivalent inactivated vaccine ( TIV from GSK ) , containing the H1N1pdm09 strain during the 2010–2011 season . Blood samples were collected before ( t0 ) and 4 weeks ( t28 ) post-vaccination . Subjects have not received the H1N1 vaccine or had flu-like symptoms at the time of enrollment . Immune responses to the H1N1pdm09 vaccine in young and elderly individuals at t0 and t28 were measured by HI assay , as previously described [6] [33] . PBMC were collected by density gradient centrifugation on Ficoll–Paque premium solution ( GE Healthcare ) . Cells were then washed three times with PBS and frozen . Frozen PBMC were thawed in a 37°C water bath and washed twice with medium ( RPMI 1640 ) , resuspended , rested for 1 h , and then counted in trypan blue to evaluate cell viability . PBMC ( 106/ml; ≥80% viability ) were cultured in complete medium ( RPMI 1640 , supplemented with 10% FCS , 10 µg/ml Pen-Strep , 1 mM Sodium Pyruvate , 2×10−5 M 2-ME and 2 mM L-glutamine ) . PBMCs from individuals in the 2009–10 season were stimulated with H1N1pdm09 monovalent vaccine ( 2 µl/106 cells ) . PBMC from the individuals in the 2010–11 season were stimulated with the H1N1pdm09 egg-derived mono-bulk subunit antigens ( A/California/07/2009; Novartis Vaccines and Diagnostics , Siena , Italy ) at the same concentration . Although B cells in the PBMC cultures have been stimulated in the presence of other cell types ( primarily T cells and monocytes ) , our endpoint is to measure a B-cell response and AID is expressed exclusively in B cells . In previous studies , we have compared the response of purified B cells and PBMC cultures and demonstrated that the AID mRNA fold-increase in response to vaccine stimulation in both in vitro cell cultures was comparable [7] . At the end of stimulation , cells were harvested and mRNA extracted using the μMACS mRNA isolation kit ( Miltenyi Biotech ) according to the manufacturer's protocol , eluted into 75 µl of elution buffer , and stored at −80°C until use . qPCR was performed as previously described [5] . We always assay the same number of B cells in PBMC cultures of young and elderly individuals . This was confirmed by measuring B cell subsets ( numbers and frequencies ) using flow cytometry . Furthermore , we measure the fold-increase from day 0 ( t0 ) to day 28 ( t28 ) of the AID response generated in cultures , since the numbers of B cells are maintained between these two time points as determined by flow cytometry . Moreover , we have shown that the subsets which make AID in vitro are the naïve and IgM memory B cells [4] , [14] , which are present in similar percentages in young and elderly individuals [5] . Reactions were performed with Taqman Master mix and primers , as described [4] , [6] . The number of cycles at which transcripts reached a cycle threshold ( Ct ) for AID and GAPDH as control was determined and used to calculate ΔCt ( target gene cycles relative to GAPDH cycles ) . A two-fold rise in AID mRNA , indicates a positive response to the vaccine with the t28/t0 values reporting the fold rise in AID mRNA [4] [5] , [6] . We have conducted a series of validation experiments including titration ( 5-fold serial dilutions ) of cDNA used as template in the qPCR assay , and found that the ratio of AID/GAPDH does not change irrespective of the cDNA template used in the qPCR . We validate the qPCR conditions at the beginning of each flu season using the relevant vaccine strains . We also report the AID mRNA values at t0 , which reflect the functional baseline level for each subject at the time of vaccination . Steady-state equilibrium binding of pre- and post-H1N1 human vaccine plasma polyclonal antibodies was monitored at 25°C using a ProteOn SPR biosensor ( BioRad ) . In the SPR assays described in this study , we are only measuring the antibody off-rate constants which describe the stability of the complex . The fraction of antigen-antibody complexes decaying per second were determined directly from plasma sample interaction with properly folded , H1N1pdm09 functional HA1 globular domain and HA2 stalk domain proteins [34] during the dissociation phase . We have previously demonstrated that the antigen-antibody dissociation kinetics were independent of the antibody concentration in the human polyclonal sera/plasma . In these SPR assays serial dilutions of the polyclonal sera ( 10 , 33 , 100 ) were run on the HA coated protein chip surface . The kinetics of dissociation were identical as indicated by parallel lines in the dissociation phase and hence independent of serum antibody concentration [9] , [10] . Antibody off-rate constants were calculated using the BioRad ProteOn manager software for the heterogeneous sample model as previously described [9] . To improve measurements , the off-rate constants were determined from two independent SPR runs . | Antibody affinity maturation is a key aspect of an effective immune response to vaccines , likely to have an impact on clinical outcome following exposure to pathogens . Activation-Induced Cytidine Deaminase ( AID ) in B cells is a key enzyme involved in antibody class switching and somatic hypermutation , required for antibody affinity maturation . This human study demonstrated for the first time that induction of AID following H1N1pdm09 influenza vaccination directly correlated with in-vivo antibody affinity maturation against the hemagglutinin globular domain ( HA1 ) , containing most of the protective targets . Importantly , age differences were found . In younger adults , significant affinity maturation to the HA1 globular domain was observed , which associated with higher initial levels of AID and >2-fold-increase in AID after vaccination . With increased age , a drop in AID activity post-vaccination correlated with lower affinity maturation of the polyclonal antibody responses against the pandemic influenza HA1 . However , in a subset of elderly ( >65 yr ) , high affinity antibodies against the HA1 were present prior to vaccination but , in the absence of AID , did not undergo further maturation . Therefore , vaccination of divergent human populations , especially older individuals , should take into consideration their individual AID status and the history of exposure and vaccination against the specific pathogen . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"humoral",
"immunity",
"medicine",
"viral",
"vaccines",
"influenza",
"immunology",
"microbiology",
"vaccines",
"vaccination",
"infectious",
"diseases",
"biology",
"immune",
"response",
"clinical",
"immunology",
"immunity",
"virology",
"viral",
"diseases"
] | 2012 | AID Activity in B Cells Strongly Correlates with Polyclonal Antibody Affinity Maturation in-vivo Following Pandemic 2009-H1N1 Vaccination in Humans |
Devil facial tumour disease ( DFTD ) is a fatal , transmissible malignancy that threatens the world's largest marsupial carnivore , the Tasmanian devil , with extinction . First recognised in 1996 , DFTD has had a catastrophic effect on wild devil numbers , and intense research efforts to understand and contain the disease have since demonstrated that the tumour is a clonal cell line transmitted by allograft . We used chromosome painting and gene mapping to deconstruct the DFTD karyotype and determine the chromosome and gene rearrangements involved in carcinogenesis . Chromosome painting on three different DFTD tumour strains determined the origins of marker chromosomes and provided a general overview of the rearrangement in DFTD karyotypes . Mapping of 105 BAC clones by fluorescence in situ hybridisation provided a finer level of resolution of genome rearrangements in DFTD strains . Our findings demonstrate that only limited regions of the genome , mainly chromosomes 1 and X , are rearranged in DFTD . Regions rearranged in DFTD are also highly rearranged between different marsupials . Differences between strains are limited , reflecting the unusually stable nature of DFTD . Finally , our detailed maps of both the devil and tumour karyotypes provide a physical framework for future genomic investigations into DFTD .
The Tasmanian devil ( Sarcophilus harrisii ) , the world's largest extant carnivorous marsupial , was recently listed as an endangered species , primarily due to the emergence of a fatal , transmissible cancer known as devil facial tumour disease ( DFTD ) [1] , [2] . Since the first reports of the disease in northeastern Tasmania in 1996 , DFTD has rapidly spread to over 70% of the devil's range , causing population declines of around 90% in some regions [2] , [3] . DFTD could lead to extinction of the species in the wild within 25–35 years [4] . Devils have a life expectancy of approximately six months from the first appearance of a lesion , with death occurring due to starvation , secondary infections and metastases [5] . In the absence of a vaccine or treatment for the disease , current measures taken to conserve the devil are focussed on breeding disease-free insurance populations in captivity [6] . Obviously , a deeper understanding of DFTD pathogenesis is required in order to help conserve this iconic species . A striking feature of DFTD is that the tumour is a clonally derived cell line transmitted as an allograft between individuals by biting [7] , [8] , [9] . The only other example of a contagious cancer in the wild is Canine Transmissible Venereal Tumour ( CTVT ) in dogs , a histiocytic tumour typically transmitted through coitus [10] . Cytogenetic analysis of DFTD tumours from different individuals provided the first evidence of DFTD clonality . Pearse and Swift [8] demonstrated that DFTD tumours from 11 different individuals sampled from different locations in eastern Tasmania shared the same karyotype . This karyotype is highly rearranged , with loss of parts or all of three autosomes and the addition of four marker chromosomes . Since the description of the original DFTD karyotype , new karyotypic strains of the DFTD tumour have been identified , suggesting that the tumour is evolving [5] . G-banding shows that these strains are closely related , as would be expected of derivations of the original karyotype [5] . Determining the genome arrangement of normal devil chromosomes is an essential first step to characterising the rearrangements that have occurred in DFTD , identifying the genes that may have been altered by rearrangement and assessing how the tumour is evolving . However , prior to the emergence of DFTD very little cytogenetic analysis had been carried out on this species , and no molecular or mapping work . The devil diploid karyotype consists of six pairs of autosomes and a pair of sex chromosomes ( XX in females and XY in males ) . The devil belongs to the Family Dasyuridae , a group of marsupials renowned for their highly conserved 2n = 14 karyotypes [11] , [12] , [13] . Cross-species chromosome painting revealed homologous chromosome segments amongst even the most distantly related marsupials , including two dasyurid species , Sminthopsis macroura [14] and Sminthopsis crassicaudata [15] . These regions of homology can be applied to the devil karyotype enabling comparison with mapped and sequenced marsupial genomes , such as the South American opossum ( Monodelphis domestica ) [16] and the tammar wallaby ( Macropus eugenii ) [17] , and allowing us to predict which genes will be present on each devil chromosome . The origins of the marker chromosomes and the extent of rearrangement of tumour chromosomes are of intense interest . Initial G-band analysis of DFTD tumour chromosomes [8] showed that both copies of chromosome 1 and part of one copy of chromosome 2 ( mislabelled as chromosomes 2 and 1 respectively in Pearse and Swift [8] ) , as well as both copies of chromosome 6 were replaced by marker chromosomes . However , this method is unable to detect molecular homology of marker chromosomes and lacks the resolution to determine the extent of rearrangement within the tumour karyotype . Chromosome painting of the tumour using whole chromosome probes generated from normal devil chromosomes , provides molecular information on the gross homologies between normal and DFTD chromosomes , although it cannot detect internal rearrangements . Gene mapping of normal and tumour cells provides information on changes in gene order , and detects rearrangements at a much higher resolution . We have therefore employed these two complementary approaches to identify the origins of the DFTD tumour marker chromosomes and determine the extent of rearrangement between normal and DFTD chromosomes . We used chromosome painting to identify large regions of homology between normal and DFTD chromosomes . We mapped genes from the ends of opossum-wallaby evolutionary conserved gene blocks to identify chromosome homology on a finer scale . This allowed us to determine the origin of marker chromosomes and evaluate the differences between emerging strains of the disease . Our findings demonstrate the unusually stable nature of the tumour karyotype , even between strains , point to candidate genes involved in tumourigenesis and indicate that certain regions of the genome are hotspots for rearrangement in marsupial evolution and in DFTD tumours . This information provides a framework for studies of genome changes at the sequence level that underlie the transmissible tumour in the Tasmanian devil .
To efficiently construct the physical map , conserved blocks of genes were identified by comparing the anchored opossum genome assembly to the physical map of the tammar wallaby genome . A total of 60 opossum/wallaby conserved gene blocks were identified , covering the entire genome . Genes located near the ends of these conserved gene blocks were used to search the devil transcriptome [7] and overgos were designed using devil sequence to isolate BACs for these genes from a male 6 . 5× genome coverage devil BAC library ( VMRC-50 ) with an average insert size of approximately 140 kb . These BACs were mapped onto normal devil male metaphase chromosomes . Smaller conserved gene blocks ( estimated from the opossum genome to be less than 4 Mb ) were localized by mapping a single gene , rather than genes from either end of the block , due to the limitations of FISH to accurately determine the orientation of BACs that are so close together . The resulting physical map contains 105 genes ( see Table S1 for a list of genes and their corresponding BACs ) . Chromosomes 1 , 3 and X are the most densely mapped . Few genes have been mapped to chromosome 5 ( Figure 1; Table 1 ) , even after new and/or redesigned overgos were used to screen the BAC library ( see Table S2 for details on overgo success rate ) . The low success rate for chromosome 5 suggests that clones from this chromosome may be under-represented in the library , perhaps due to a lack of EcoRI fragments of the size selected for library construction . Cross-species chromosome painting was used to predict which devil chromosome would contain each of the genes . Although devil chromosomes have not been used for such cross-species hybridisations , chromosome painting has been performed on two other dasyurid species ( Sminthopsis crassicaudata and S . macroura ) [14] , [15] . As the dasyurid karyotype is highly conserved , we could use data from these two species to make predictions in the devil . Of the 105 genes mapped , only four ( ARL4A , BET1L , LECT1 , SLITRK5 ) mapped to unexpected locations . Sequencing was used to confirm that these four BACs did contain the gene of interest ( Figure S1 ) . Overall , the gene mapping data correlated with cross-species chromosome painting on Sminthopsis species and extends the observation of a highly conserved dasyurid karyotype . One discrepancy between the reported painted [14] , [15] , [18] and G-banded karyotypes [19] and the karyotype described by Pearse and Swift [8] is whether chromosome 1 is the large metacentric or submetacentric chromosome . Here we used the long-established classification of Martin and Hayman [11] , [20] , which was subsequently used in classic comparisons with other marsupial karyotypes [12] and in chromosome painting studies [14] , [15] , designating chromosome 1 as the large submetacentric chromosome in dasyurids , corresponding to conserved segments C1 to C6 based on chromosome painting [15] . Chromosome 2 consists of conserved segments C7 , C8 and C9 [15] . Although chromosome painting confirms that even distantly related marsupials share large regions with DNA homology [14] , [15] , [18] , our comparison between gene arrangement in the devil , opossum and wallaby shows that within some of these blocks , gene order has been highly rearranged by multiple inversions ( Figure 2 ) . The most conserved chromosome amongst the marsupials was the long arm of devil chromosome 3 ( Figure 2A , the short arm of devil chromosome 3 corresponds to wallaby chromosome 6 and opossum chromosome 7 , Figure S2A ) , which appears as a single block conserved between the wallaby and devil , although there have been two inversions in this region with respect to the opossum . Highly rearranged chromosomes include the devil X chromosome ( Figure 2B ) and chromosome 1 . Chromosomes 2 and 4 show an intermediate level of rearrangement ( Figure S2B and S2C ) . Too few genes were mapped to chromosomes 5 and 6 to determine the extent of conservation or rearrangement between species . By mapping the ends of opossum/wallaby conserved gene blocks we hoped that we could virtually assign each gene within these conserved gene blocks to a location on devil chromosomes . The extent of rearrangement between these three species makes the construction of a virtual map based on both gene content and gene order difficult , and would require the localization of many more genes . However , we are able to predict the gene content of each block and hence , the gene content of each chromosome . Since the DFTD karyotype was first reported in 2006 , multiple karyotypic ‘strains’ have been discovered [5] . The various strains are characterized by minor cytogenetic rearrangements that demonstrate ongoing tumour evolution as the disease spreads across Tasmania . Only a small number of readily identifiable rearrangements distinguish the three strains; the basic composition of the DFTD karyotype is preserved . The random gains , losses and translocations that characterize unstable tumour karyotypes are not present in any of the DFTD strains , which are therefore considered stable . The three strains are readily identifiable with G-banding; however , this technique is insufficiently precise to determine the genomic regions that are specifically rearranged in each strain . The comparatively finer technique of chromosome painting permitted a more detailed characterisation of the DFTD karyotype , as well as the progressive chromosome changes that distinguish three tumour strains . Chromosome painting using whole chromosome probes derived from normal devil chromosomes ( see Figure S3 for the flow karyotype ) was carried out on eight tumour samples comprising Strains 1 , 2 and 3 . Samples were collected from animals at different locations throughout Tasmania ( refer to Figure S4 for strain details ) . The diagnostic DFTD karyotype is present in all tumours , with only subtle cytogenetic differences between strains . All DFTD cell lines were karyotypically stable in cell culture , with no progressive chromosome rearrangements detected after multiple ( greater than 10 ) passages . Thus the tumour karyotype was found to be remarkably stable in vivo and in vitro , with only minor cytogenetic differences between strains , a surprising result considering the rapid proliferation and malignant behaviour of neoplastic cells . Painting of cells from DFTD Strain 1 revealed that the four marker chromosomes were derived predominantly from chromosomes 1 , 5 and X ( Figure 3 ) . The giant marker chromosome ( M1 ) consisted almost entirely of chromosome 1 material , which also made smaller contributions to markers 2 and 3 ( M2 and M3 , respectively ) , as well as a small insertion in chromosome 2p . The chromosome 5 probe hybridised to the single copy of chromosome 5 present in the tumour , as well as to M2 and M4 in relatively simple rearrangements . X chromosome rearrangements were more complex , with small insertions of X material in M2 and chromosome 2p , adjacent to the chromosome 1 insertion , and extensive rearrangement between chromosomes 6q , M1 and M3 . The Y chromosome could not be detected within the tumour using a probe generated by manual microdissection ( Figure S5 ) , suggesting that the original tumour derived from a normal cell of a 2X female . Based on both G-banding and chromosome painting results , strain 1 cells were found to retain the basic DFTD karyotypic framework , whereas Strains 2 and 3 were marked by additional rearrangements . In Strain 2 and a proportion of Strain 3 tumours , an additional marker chromosome M4 was hybridized by the chromosome 4 paint throughout the long arm , and an additional reciprocal translocation between chromosomes 4 and 5 ( Figures S6 and S7 ) . These strains had an additional marker chromosome M5 , which completely hybridised to the X paint ( Figure 4 , Figure S6 ) . Strain 3 karyotypes were found to be somewhat more complicated than for Strains 1 and 2 , showing variation of painting patterns between tumour cell lines isolated from different animals , and the presence of two distinct sub-strains in two of the three tumours examined . M4 was variably present in Strain 3 tumours , with loss of this marker in 0–64% of metaphases in different tumour cell lines ( see Figure S6 ) . The variable loss of M4 was interpreted as a relatively minor change and was not considered indicative of more broad scale karyotypic instability . Strain 3 karyotypes were otherwise similar to those of Strain 2 , with the exception of chromosomes 4 and 5 , which were further rearranged in a proportion of tumours . Figure S6 catalogues the chromosome 4 and 5 rearrangements unique to Strain 3 tumours and compares the sub-strains present in two of the tumours examined . An additional translocation between chromosomes 4p and M4q was present in some cells in Strains 3A and 3C . This translocation was present in all metaphases of Strain 3A , compared with only 12 . 5% ( 1 out of 8 ) of Strain 3C metaphases , and was absent in Strain 3B . Strain 3B also exhibited some heterogeneity; 36% of cells lacked M5 ( 18 out of 50 ) , and 58% ( 29 out of 50 ) lacked M4 . In cells lacking M4 , the chromosome 5 paint hybridised to the short arm of the giant marker , replacing the X chromosome signal present at this location in all other tumours . In the 36% ( 18/50 ) of tumours that had M4 , the chromosome 5 paint hybridised to the long arm of M4 , as for Strain 1 tumours . Paints generated from flow-sorted normal devil chromosomes have therefore revealed the origin of the genomic material that comprises each marker chromosome , as well as several insertions undetectable with G-banding . Painting also demonstrated the extent to which chromosomes 1 , 4 , 5 and the X chromosome are rearranged in DFTD . None of this information could be gained from earlier G-banding studies . Our findings indicate that progressive rearrangements of chromosomes 4 , 5 and the X chromosome distinguish the three strains , and that multiple Strain 3 tumours are composed of at least two sub-strains , present in varying proportions , implying that passage of the tumour from animal to animal is usually via multiple cells . The resolution afforded by painting is insufficient to identify the genetic constitution of breakpoints associated with tumour cell rearrangements . To pinpoint rearrangements in the DFTD tumour , we therefore constructed a physical map of the three tumour strains described above , using the same 105 genes we used to construct the physical map of the normal devil genome . This map of the tumour genomes ( Figure 5 ) shows that rearrangement in the tumour has been more extensive than could be detected by chromosome painting ( Figure S8 ) . Genes from chromosome 1 were found on one copy of distal 2p , the long arm of M1 , distally on both arms of M2 and much of M3 , as was also indicated by chromosome painting . In addition , gene mapping demonstrated the presence of chromosome 1 genes on the short arm of M1 and M4 . Gene mapping also revealed an addition of at least one chromosome 3 gene ( ABCA12 ) to the long arm of one copy of chromosome 2 , and the addition of at least one chromosome 5 gene ( IPO8 ) to the long arm of chromosome 3 ( Figure 5 ) . Four of the 12 genes mapped to the short arm of chromosome 4 are found on the short arm of M2 and long arm of M4 ( e . g . SOST and PGBD2 , Figure 6A ) . Repositioning of the centromere was also detected , and reordering of many of the genes remaining on chromosome 4 in the tumour ( e . g . GNL1 and RUNX2 ) ( Figure 6B ) . Gene BET1L mapped to different locations on the two homologues of chromosome 6 and another copy of BET1L was found to be located on M3 ( Figure 6C ) . As predicted by chromosome painting , X chromosome genes were located on one homologue of 2p , one homologue of chromosome 6 , the short arm of M1 , distal M2q and proximal M3 . In addition , at least one X-borne gene ( MECP2 ) was found on the short arm of M4 ( Figure 6D ) . Both painting and mapping data identified two copies of chromosome 6 present in DFTD and one intact copy of chromosome 5 with the other copy distributed across marker chromosomes , conflicting with the original DFTD karyotype reported by Pearse and Swift [8] . Given the reshuffling of gene order and the addition of a region from the X chromosome inserted on one homologue of chromosome 6 , it is not surprising that the identity of this chromosome could not be accurately determined by G-banding . Likewise , the size difference between the two large metacentric chromosomes was initially interpreted as a deletion of part of the long arm on one homologue . However , our gene mapping shows that the size differences between the two copies of chromosome 2 are due to addition to the short arm of one homologue of regions bearing genes from chromosomes X and 1 . Confirming our results from chromosome painting , gene mapping revealed only subtle differences between tumour strains . The additional marker chromosome ( M5 ) of Strain 2 was found to contain one gene from the X chromosome ( MECP2 ) and one gene from chromosome 1 ( SHARPIN ) . The only other detectable difference between Strains 1 and 2 is the location of X chromosome genes HEPH and THO2C , which were observed to be near to each other , but not adjacent , in the normal devil genome ( Figure 7 ) . In Strain 2 they mapped to the same location on M2 , but in Strain 1 , they were found to be separated by chromosome 1 and 5 genes . A readily distinguishable difference between G-banded karyotypes of tumour strains was found to be the deletion of part of the short arm of chromosome 3 uniquely in Strain 3 . We have confirmed this by gene mapping and show that the region deleted spans from MDH1B on distal 3p to TGFBRAP1 on proximal 3p . Only one copy of the chromosome has this deletion in strain 3B , but both copies have the deletion in Strain 3A ( Figure 7 ) and no signals were observed for these genes on any other chromosome , suggesting these genes are completely absent from the tumour genome . The deletions appeared to be the same on both copies of chromosome 3 , suggesting that the normal member of the pair may have been lost , and the deleted copy reduplicated . The three Strain 3 tumours also have variations in the arrangement of chromosome 4 and 5 genes ( Figure 7 ) . Genes from the short arm of chromosome 4 were observed to be absent from one copy of the chromosome in Strain 3A , and this deletion is also present in 20% of Strain 3C metaphase spreads . In addition , Strain 3B was found to have retained TPST1and SENP2 on chromosome 4 ( these genes were found on M4 in all other strains ) , although SENP2 was observed to be translocated to 4q . This strain was shown also to have acquired an additional copy of C17orf101 on the short arm of M2 . Strain 3C had three copies of ST6GALNAC5 , one copy on each of the chromosome 4 homologues observed in all strains , as well as an additional copy on the short arm of M2 . In all three Strain 3s , chromosome 5 genes were detected on the short arm of M2 , and in strain 3C also on M5 . Gene mapping can also detect variation in the numbers of copies of a gene , revealing a copy number increase or deletion of small regions of the genome that are hard to detect by chromosome painting . Nearly all genes mapped in the tumour were observed to be present in two copies , but for Strain 1 , we identified twelve autosomal genes present in only one copy and three autosomal genes present in three copies ( Table 2 ) . Significantly , we found that 11 of the 14 genes from the X chromosome were present in two copies , consistent with the origin of the original tumour from an XX female .
The transformation of a normal cell to a cancerous one involves the accumulation of mutations , often in tumour suppressor genes or oncogenes . There is a growing list of such genes perturbed in human cancers , making it difficult to know where to begin searching for candidate genes involved in tumourigenesis in DFTD . Our mapping data allows us to predict where many of the most common tumour suppressor genes and oncogenes are located in DFTD , and whether these sites are located in regions of the devil genome that were rearranged in the tumour . From the list of common cancer genes ( Table 3 ) , we find that a large number are located on devil chromosome 1 . Significantly , this chromosome has undergone extensive rearrangement in the tumour ( Figure 8A ) . Several genes ( APC , MYC , NF2 , MLH1 ) stand out as potentially playing a role in DFTD tumourigenesis , being predicted to be close to genes that have one copy deleted in DFTD ( REEP5 , ENM01188 , OSBP2 , WDR48 respectively ) and hence , they themselves may be perturbed . The Schwann cell origin of DFTD [7] makes the tumour suppressor NF2 a particularly interesting gene to examine more closely in future studies . In humans , loss of NF2 function is linked to tumours of the central nervous system , particularly benign tumours such as schwannomas [23] , although in mice loss of NF2 has been associated with a variety of malignant tumours [24] . We predict , based on the opossum genome assembly , that NF2 is approximately 2 Mb away from the mapped gene OSBP2 , a gene that maps to only one position on the short arm of chromosome 2 in DFTD ( Figure 8A ) . Devil facial tumour disease is a rare exception to established models of tumour development and progression , as demonstrated by cytogenetic evidence [8] . The classic model of stepwise carcinogenesis describes a gradual process in which neoplastic cells progress through a spectrum of increasingly malignant phenotypic changes that correlate with escalating genomic chaos [25] . This is best exemplified by human colorectal tumours , in which the transformation of benign dysplastic lesions into invasive carcinomas is associated with an accumulation of gross cytogenetic aberrations [26] , [27] . Randomly acquired genetic mutations that afford neoplastic cells a competitive advantage are propagated in waves of clonal expansion so that increasingly malignant cells are selected for in a process akin to Darwinian evolution . By contrast , the cancer stem cell ( CSC ) model posits that only a proportion of neoplastic cells have the capacity for self-renewal and tumour initiation , and these cells are the drivers of malignancy [28] . These two theories are neither conflicting nor mutually exclusive , and both account for the intra-tumoral heterogeneity typically present in solid and hematologic malignancies . In contrast , DFTD is a stable , clonal cell line transmitted from animal to animal by biting . Its biological behaviour within wild devil populations renders it a somatic cell pathogen that forms proliferative masses upon transplantation . A lack of genetic diversity between animals at functionally important MHC loci [9] and the epidemiologic dynamics of DFTD transmission [29] set the stage for the devastating disease outbreak that now threatens extinction of Tasmanian devils . The genomic events that underpinned the formation of the original devil tumour are uncertain; however , our chromosome painting and BAC mapping results have pinpointed candidate genes and elucidated the gross cytogenetic restructuring that produced the original tumour and switched a Schwann cell in a single sentinel animal into the pathway to carcinogenesis . Consistent with previous G-banding and genotyping results [7] , [8] , [9] our chromosome painting experiments support the hypothesis that DFTD derived from a clonal cell line in 1996 . The absence of Y-chromosome sequences ( Figure S5 ) suggests that the sentinel animal that harboured the original tumour was female . The presence of two copies of 11 out of 14 X-borne genes supports this hypothesis . It is possible that the neoplastic cell that ultimately became transmissible was a clonal stem cell ( CSC ) . This is consistent with the limited heterogenetiy of neoplastic cells , their poorly differentiated morphology [30] and their gene expression profile [7] . Our observation of limited divergence into several strains and sub-strains implies that the basal tumour karyotype was established early in tumour evolution , and has remained extraordinarily stable over the subsequent fifteen years . Thus an alternative hypothesis is that all tumour strains are the same age and represent various subclones of an original , heterogenous tumour in the sentinel animal . However , subclones must have been all capable of self-renewal and tumour initiation , which seems rather unlikely as few cells independently acquire properties of CSCs . A third , intriguing , possibility is that the DFTD karyotype was generated in a single episode of massive genomic restructuring . Termed chromothripsis , this phenomenon was recently described in a variety of solid and hematologic malignancies [31] . The genomic signature of chromothripsis is typified by complex remodelling of a small number of chromosomes with minimal loss of heterozygosity and variation in gene copy number . It is clear that complex chromosome rearrangements in DFTD are localised to well demarcated genomic regions . BAC mapping results demonstrate that chromosomes 1 and X are particularly fragmented , with dozens of DNA breaks and fusions contained to only a small portion of the genome . Our observation that chromosome 1 has undergone the same numerous rearrangements in all strains suggests that rearrangement of this chromosome as a result of chromothripsis was the initial step in the development of DFTD . Stephens et al [31] suggest that chromothripsis occurs when cells undergo catastrophic chromosome rearrangements , during which well delineated regions of the genome are reduced to tens or hundreds of fragments that are haphazardly fused by nonhomologous end-joining DNA repair machinery . What might incite such dramatic genomic restructuring is unknown , though the authors suggest that breakage-fusion-bridge ( BFB ) cycles associated with telomere loss could cause the catastrophic genomic restructuring of chromothripsis . This is a particularly intriguing speculation , as telomere length varies between chromosomes; those chromosomes with the shortest telomeres are predisposed to telomeric fusions and are consequently drivers of BFB cycles and chromosome rearrangement [32] . The DFTD karyotype may be a snapshot of a brief period of localised genomic instability associated with focal telomere attrition , eventually rescued by recruitment of telomerase expression . The clonal passaging of DFTD from animal to animal over a protracted period provides a unique opportunity to study the long-term karyotype evolution of a solid tumour . Surprisingly , we found that cytogenetic differences between tumour strains are minimal . The eight DFTD cell lines examined in this study were established from primary lesions in male and female devils trapped in various locations throughout Tasmania over a period of three years ( Figure S2 ) . We found both inter-strain and intra-strain differences of similar magnitude , highlighting the stability of the DFTD genome while suggesting that karyotype evolution continues . Additionally , the presence of multiple sub-strains suggests that upon transmission , the tumour inoculum contains mixtures of cell lines that may have diverged over some years . For instance , the two 3B sub-strains are distinguished by the variable loss of marker chromosome M5 , subtle variations in chromosome 5 rearrangements and the absence of an additional chromosome 4 rearrangement that marks other Strain 3 tumours . The differences within this tumour are more complex than the subtle rearrangements that distinguish Strains 1 and 2 . This observed pattern of intra-tumour chromosome variability is consistent with observations that the tumour is passed from animal to animal by biting , during which many clumps of tumour cells are dislodged from the mouth of the affected animal [33] . The long-term stability of tumour chromosomes , both in vivo and in vitro , indicates that DFTD does not share the overt genomic instability typical of many solid tumours in humans and mice . Nevertheless , the predominance of chromosome 4 , 5 and X permutations among and within strains may correlate with mild chromosome instability localised to these chromosomes . Perhaps selection is acting on the DFTD karyotype to maintain the tumourigenic properties of a DFTD cell , while tolerating genomic instability in regions of the genome not essential for survival of a DFTD cell . This is consistent with the hypothesis that chromosome 1 rearrangement was the initial step in the development of DFTD and that the maintenance of these rearranged chromosome 1 regions is critical for the survival of DFTD in the devil population . Conversely , continued perturbations of chromosomes 4 , 5 and X are neutral , having no affect on DFTD tumourigenesis . There are no data that attaches any clinical significance to the karyotypic strains , nor is it known whether the emergence of new karyotypic strains correlates with meaningful phenotypic changes . The provision of detailed descriptions of strain karyotypes will make it possible to investigate this important question in more depth . It appears that certain regions of the human genome are ‘hotspots’ for rearrangement in tumours [34] and there has been much debate about whether these regions are the same parts of the genome that display the most rearrangement when comparisons of gene arrangement are made between eutherian mammals . Cancer-associated breakpoints in humans have been frequently reported to co-localise with evolutionary breakpoints , regions in which chromosomal breaks have occurred more than once during eutherian evolution [34] . However , a more recent study which localised breakpoints on a much finer scale refuted this claim by finding no evidence of more frequent co-localisation of evolutionary and cancer breakpoints [35] . Perhaps evolutionary and tumour breakpoints do not occur at exactly the same base pair position in the genome , but are concentrated in specific regions of the genome that are more susceptible to breakage , both during the course of evolution , and tumourigenesis . Intriguingly , the chromosomes most rearranged in DFTD tumour lines are the same ones that are most rearranged between devil , wallaby and opossum genomes . Chromosome 1 is a good example , since there has been extensive rearrangement of this chromosome in DFTD and between different marsupial species ( Figure 8 ) . Furthermore , the same parts of this chromosome are less or more subject to rearrangement both in the tumour and between species . The region from EFCAB1 to KCTD1 on the long arm of chromosome 1 is intact ( conserved in gene order ) on both marker chromosomes M1 and M2 , and is conserved ( gene order ) as a block in wallaby and devil , suggesting that this region has been less susceptible to rearrangement in DFTD and during marsupial evolution . The remainder of chromosome 1 is highly rearranged in DFTD , being spread across five chromosomes and with eight out of 12 genes present in only a single copy and one gene mapping to three different locations ( Figure 8A ) . This region has undergone extensive reshuffling between devil , wallaby and opossum ( Figure 8B ) . Regions of the genome that are relatively well conserved between species ( e . g . the long arm of devil chromosome 3 , see Figure 2A ) have remained unchanged in DFTD . Genome sequence data is required to determine whether there are sequence features in common between regions susceptible to rearrangement . The emergence of DFTD has had a disastrous effect on wild Tasmanian devil numbers , and with the devil now perilously close to extinction , intense research efforts to understand and intercept DFTD pathogenesis proceed apace . Here we contribute a detailed map of the global chromosome restructuring and intricate gene rearrangements that characterise DFTD . We provide further confirmation of the clonal transmission of DFTD and tentatively identify the sentinel animal as a female devil . Our observation that only limited regions of the genome are highly rearranged suggest that chromothripsis was the mechanism of the original tumorigenesis , and , once remodelled , the tumour karyotype has been remarkably stable during its clonal transmission from animal to animal . By anchoring genes to a reference and tumour maps , we can predict the locations of common tumour suppressor genes and oncogenes . By characterising multiple strains and sub-strains we have demonstrated the stability of the tumour genome . This study provides an important framework for future genomic studies into DFTD .
The collection of samples from devils was approved by the Australian National University Animal Experimentation Ethics Committee ( AEECP R . CG . 11 . 06 ) . Tissue samples for tumour culture were obtained from biopsies of live , wild caught Tasmanian devils and from necropsy specimens . Wild Tasmanian devils were trapped for the purposes of disease surveillance and epidemiologic studies , and were biopsied under general anaesthesia . DFTD-affected animals that were euthanased , either for humane reasons or because they were trapped in disease exclusion sites [36] , were necropsied in the field or at the Tasmanian Animal Health Laboratory . Samples were sourced from a variety of geographic locations in order to obtain representative cultures of each of the three tumour strains . Primary tumour cultures were initiated according to the Pearse and Swift [8] protocol . Briefly , tumour biopsies were washed in 10 mL Dulbecco's phosphate buffered saline ( PBS ) ( Invitrogen , Mulgrave , VIC , Australia ) supplemented with 0 . 1 mL penicillin-streptomycin solution ( Sigma-Aldrich , Castle Hill , NSW , Australia ) . Cultures were established by manually disaggregating tumour tissue using a scalpel , followed by re-suspension in 8 mL GIBCO AmnioMAX-C100 ( Invitrogen ) . Cultures were incubated at 35°C in 5% CO2 and harvested after 24 to 48 hours for diagnostic purposes and to ensure that additional chromosome rearrangements did not occur in subsequent passages . Metaphase chromosomes were prepared from a normal male devil cell line ( passage 3 ) and DFTD cultures according to standard techniques [8] . In brief , cultures were harvested after a 2 hour synchronisation with colcemid ( 10 µg/mL ) by incubating in 37°C , hypotonic solution ( 0 . 075 mM KCL ) for 18 minutes and fixation with chilled methanol∶acetic acid ( 3∶1 ) . Cell suspensions were dropped on to slides , air-dried and stored for 24 hours prior to hybridisation . A panel of six chromosome paints comprising all autosomes and the X chromosome were hybridised to metaphase chromosomes from each of the three tumour strains . Chromosome paints for devil chromosomes 1 to 6 and the X were generated from flow sorted S . harrisii chromosomes as previously described [37] . The Y chromosome paint was produced by manual microdissection of metaphase chromosomes , freshly dropped onto glass coverslips and collected with a glass needle mounted on a Ziess Axiovert I microscope [38] . Primary degenerate oligo-primed ( DOP ) PCR products were labelled with biotin-dUTP or digoxygenin-dUTP ( Roche Diagnostics , Basel , Switzerland ) in subsequent amplifications by DOP-PCR with 6MW primer ( 5′-CCG ACT CGA GNN NNN NAT GTG G-3′ ) [39] . The labelled PCR product was co-precipitated with Cot-1 DNA ( 5 ug/slide ) for suppression [40] , suspended in 15 µl of pre-warmed hybridisation buffer ( 50% formamide , 2× SSC , 10% dextran sulfate ) and denatured at 70°C for 10 min and pre-annealed for 20 min at 37°C . Metaphase spreads were denatured for 40 seconds in a 70% formamide solution at 70°C and hybridised overnight at 37°C . Post hydridisation washes were performed according to Alsop et al [40] . Biotin and digoxygenin-labelled probes were detected with avidin-FITC ( Vector Laboratories Inc . , Burlingame , CA , USA ) and anti-digoxygenin-Cy3 ( Roche Diagnostics ) , respectively . DAPI ( 4′ , 6-diamidino-2-phenylindole ) was used as a counterstain and slides were mounted in Vectashield ( Vector Laboratories Inc . , Burlingame , CA , USA ) . A Zeiss Axioplan2 epifluorescence microscope was used to visualise fluorescent signals which were captured with a SPOT RT Monochrome charged-couple device camera ( Diagnostic Instruments Inc . , Sterling Heights , MI , USA ) and processed using IP Lab imaging software ( Scanalytics Inc , Fairfax , VA , USA ) . A bacterial artificial chromosome ( BAC ) library , designated VMRC-50 , was produced using the detailed procedures of library construction described previously [41] , [42] . This library was constructed from genomic DNA extracted from the liver of a deceased two-year-old male devil ( Accession Number 08/0134 ) that was originally from Bangor , Tasmania and euthanized in 2008 due to multiple DFTD lesions and metastases to the lungs . Quality of the DNA was checked by running a pulsed field gel electrophoresis ( PFGE ) on a CHEF -DR III system ( BioRad , Hercules , CA , USA ) . The DNA was partially digested in an EcoRI/EcoRI-methylase competition reaction and size fractionated by analytical PFGE on a CHEF Mapper XA system ( BioRad ) . DNA fragments from the appropriate size fraction were ligated into the CopyControl pCC1BAC vector from Epicentre Technologies and transformed into ElectroMAX DH10B T1 Phage-Resistant E . coli cells ( Invitrogen ) . Transformants were arrayed into 384-well LB/chloramphenicol/glycerin microtiter plates ( Genetix , San Jose , CA , USA ) using colony-picking robot ( Norgren Systems , Fairlea , WV , USA ) and subsequently gridded onto 22×22 cm high-density nylon filters with a Total Array System ( BioRobotics Ltd . , Woburn , MA , USA ) . Genes located near the ends of opossum-wallaby conserved gene blocks were identified by comparing the arrangement of genes between the anchored opossum genome sequence [16] and physical map of the wallaby genome ( Deakin et al , in preparation; [43] , [44] ) . Opossum orthologues for genes located near the ends of these blocks were found in the Ensembl gene build ( MonDom5 ) and used to search the available devil transcriptome sequence [7] with BLASTN . Devil-specific overgos were designed using the Overgo Maker program ( http://genome . wustl . edu/software/overgo_maker ) using the devil orthologous sequence as the input sequence . Specificity of the resulting 40 bp probe was confirmed by BLASTN searches of the devil transcriptome , as well as the wallaby and opossum genome assemblies . Proposed overgos matching numerous positions in the wallaby and opossum genomes or many contigs in the devil transcriptome were discarded in order to avoid the detection of paralogous genes . A complete list of the overgos used in this study is provided ( Table S1 ) . BAC library filters were screened with pools of up to 60 radioactively labelled overgo pairs using the protocol described by Ross et al [45] . Dot blots were performed as described by Deakin et al [43] on the resulting positive BACs in order to determine which BACs were positive for each gene . BACs mapping to different chromosomes than predicted were subjected to direct sequencing , using an overgo as a sequencing primer according to the previously described protocol [43] . DNA from each BAC clone was isolated using the WIZARD Plus SV Minipreps DNA Purification System ( Promega , Alexandria , NSW , Australia ) , and approximately 1 µg of DNA was labelled by nick translation with either SpectrumOrange dUTP or SpectrumGreen dUTP ( Abbott Molecular Inc . , Des Plaines , IL , USA ) . Labelled probes were hybridised overnight to normal male devil or DFTD tumour metaphase chromosomes following the protocol detailed in Alsop et al [40] with one exception . Denaturation time for normal male chromosomes was 1 min 40 but , as the tumour chromosomes were observed to be more susceptible to overdenaturing , the denaturing time was reduced to 1 min for DFTD tumour chromosomes . Unbound probe was washed off slides with one wash of 0 . 4×SSC with 0 . 3% ( v/v ) Tween 20 for 2 min at 60°C , followed by a wash at room temperature in 2×SSC with 0 . 1% ( v/v ) Tween 20 for 5 sec to 1 min . Chromosomes were counterstained in DAPI and mounted with Vectashield ( Vector Laboratories Inc . ) . Fluorescent signals were visualised using a Zeiss Axioplan2 epifluorescence microscope . Images of both DAPI stained chromosomes and fluorescent signals were captured on a SPOT RT Monochrome CCD charge-coupled device camera ( Diagnostic Instruments Inc . ) and merged using IP Lab imaging software ( Scanalytics Inc ) . | The world's largest carnivorous marsupial , the Tasmanian devil , is threatened with extinction due to the emergence of devil facial tumour disease ( DFTD ) , a fatal transmissible tumour . Critical loss of genetic diversity has rendered the devil vulnerable to transmission of tumour cells by grafting or transplanting the cells while biting and jaw wrestling . Initial studies of DFTD tumours revealed rearrangements among tumour chromosomes , with several missing chromosomes and four additional marker chromosomes of unknown origin . Since then , new strains of the disease have emerged and appear to be derived from the original strain . With no prior information available regarding the location of genes on normal devil chromosomes , a necessary first step towards characterisation of chromosome rearrangements in DFTD was to construct a map of the normal devil genome . This enabled us to elucidate the chromosome rearrangements in three DFTD strains . In doing so we determined the origin of the marker chromosomes and compared the three strains to determine which areas of the genome are involved in ongoing tumour evolution . Interestingly , rearrangements between strains are limited to particular genomic regions , demonstrating the unusual stability of this unique cancer . This study is therefore an important first step towards understanding the genetics of DFTD . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"biology",
"genomics",
"cytogenetics",
"genetics",
"and",
"genomics"
] | 2012 | Genomic Restructuring in the Tasmanian Devil Facial Tumour: Chromosome Painting and Gene Mapping Provide Clues to Evolution of a Transmissible Tumour |
We aimed to identify genetic variants associated with cortical bone thickness ( CBT ) and bone mineral density ( BMD ) by performing two separate genome-wide association study ( GWAS ) meta-analyses for CBT in 3 cohorts comprising 5 , 878 European subjects and for BMD in 5 cohorts comprising 5 , 672 individuals . We then assessed selected single-nucleotide polymorphisms ( SNPs ) for osteoporotic fracture in 2 , 023 cases and 3 , 740 controls . Association with CBT and forearm BMD was tested for ∼2 . 5 million SNPs in each cohort separately , and results were meta-analyzed using fixed effect meta-analysis . We identified a missense SNP ( Thr>Ile; rs2707466 ) located in the WNT16 gene ( 7q31 ) , associated with CBT ( effect size of −0 . 11 standard deviations [SD] per C allele , P = 6 . 2×10−9 ) . This SNP , as well as another nonsynonymous SNP rs2908004 ( Gly>Arg ) , also had genome-wide significant association with forearm BMD ( −0 . 14 SD per C allele , P = 2 . 3×10−12 , and −0 . 16 SD per G allele , P = 1 . 2×10−15 , respectively ) . Four genome-wide significant SNPs arising from BMD meta-analysis were tested for association with forearm fracture . SNP rs7776725 in FAM3C , a gene adjacent to WNT16 , was associated with a genome-wide significant increased risk of forearm fracture ( OR = 1 . 33 , P = 7 . 3×10−9 ) , with genome-wide suggestive signals from the two missense variants in WNT16 ( rs2908004: OR = 1 . 22 , P = 4 . 9×10−6 and rs2707466: OR = 1 . 22 , P = 7 . 2×10−6 ) . We next generated a homozygous mouse with targeted disruption of Wnt16 . Female Wnt16−/− mice had 27% ( P<0 . 001 ) thinner cortical bones at the femur midshaft , and bone strength measures were reduced between 43%–61% ( 6 . 5×10−13<P<5 . 9×10−4 ) at both femur and tibia , compared with their wild-type littermates . Natural variation in humans and targeted disruption in mice demonstrate that WNT16 is an important determinant of CBT , BMD , bone strength , and risk of fracture .
Osteoporosis is a common skeletal disease characterized by reduced areal bone mineral density ( BMD ) and defects in the microarchitecture of bone , resulting in an increased risk of fragility fracture [1] . Osteoporotic fractures affect between one third to one half of white women [2] and currently incur direct costs exceeding $19 billion per year in the United States alone [3]; and this socio-economic burden is increasing with the ageing of industrial societies [4] . Twin and family studies have revealed that genetic factors can explain up to 85% of the variation in peak BMD [5] , [6] . Since 2007 , we and others have published several genome-wide association studies ( GWAS ) for osteoporosis and related traits [7] , [8] , [9] , [10] , [11] , [12] , [13] , [14] identifying multiple common variants associated with BMD and highlighting biologic pathways that influence BMD . Most osteoporotic fractures occur at peripheral sites , mainly containing cortical bone , after the age of 65 [15] . As indicated by a recent study , bone loss at this age is mainly due to loss in cortical and not trabecular bone [16] . In human cadaver femurs , cortical bone has been reported to be the main determinant of the femoral neck bone strength , while trabecular bone only contributes marginally to bone strength at this site [17] . Evidence implicating cortical thinning as a risk factor for hip fracture has also been presented [18] . The heritability for cortical thickness , measured using computed tomography , has been reported to be as high as 51% [19] . BMD is a complex trait , obtained from a 2-dimensional projectional scan of the given bone with dual x-ray absorptiometry ( DXA ) . Although BMD is the most clinical useful measure for diagnosing bone fragility ( osteoporosis ) , it fails to provide a detailed skeletal phenotype necessary to discern traits such as bone geometry and volumetric BMD ( vBMD ) [20] . Most of the loci or genes identified have been associated with BMD at lumbar spine and/or femoral neck , sites rich in trabecular bone . Therefore , we hypothesized that investigating BMD at the forearm , a primarily cortical bone site , as well cortical bone thickness , a trait with high heritability , would serve as successful strategies to identify novel bone related genetic loci . Forearm fractures are among the most common fractures , affecting 1 . 7 million individuals per year . In contrast to hip fractures [21] , forearm fractures have been shown to be highly heritable , with estimates of 54% [22] . To our knowledge , no GWA studies for cortical bone thickness , forearm BMD or fractures have been published . Importantly , we are aware of only one previous locus [12] that has been associated with risk of fracture even in large-scale meta-analytic efforts at a genome-wide significant level ( reviewed previously ) [23] , [24] , [25] . In this study , we performed two separate GWA meta-analyses in order to identify loci for cortical bone thickness of tibial diaphysis and BMD at the distal radius . Firstly , we performed a GWA study of three large and well-characterized independent discovery cohorts of 5 , 878 samples with the aim of identifying genetic loci for cortical thickness . SNPs meeting GWAS significance in the discovery meta-analysis were also tested for association in a large replication cohort ( N = 1032 ) . In the second and separate GWA meta-analysis , we combined genome-wide association results of 5 , 672 samples with BMD measurement at the forearm site from five cohorts; we then sought evidence of association of selected genome-wide significant signals in three cohorts comprising 5 , 763 individuals for forearm fracture . To determine the possible functional role of the identified genes on cortical bone thickness and bone strength , we generated mice with inactivated genes and investigated their skeletal phenotype . The resultant findings increase our understanding of the genetic basis of osteoporosis and osteoporotic fracture .
Anthropometrics , and bone variables for the three discovery GWAS cohorts and one replication cohort are presented in Table S1 . Marked deviation from the null distribution amongst the lowest observed p-values were observed for the meta-analysis results ( Figure S1 ) . The results showed that the greatest evidence for association between genetic variation and tibial cortical thickness was seen for rs9525638 on chromosome 13 , slightly upstream of the RANKL gene ( −0 . 11 standard deviations [SD] per T allele , P = 3 . 3×10−10 ) ( Table 1 , Figure S2 and Figure S3 ) . The second strongest genetic signal ( rs2707466 ) for cortical thickness was located at the WNT16 locus ( −0 . 10 SD per C allele , P = 5 . 9×10−9 ) ( Table 1 , Figure 1 and Figure S2 ) . The SNP rs2707466 represents a missense polymorphism ( Thr>Ile ) located in the fourth exon of WNT16 . We selected our top two regions , the RANKL and WNT16 loci , with SNPs with P<1×10−5 and carried out analyses conditional on the most associated SNPs in each region: rs9525638 and rs2707466 , respectively . When conditioning on the most significant SNP in the WNT16 region ( rs2707466 ) an additional suggestive signal ( rs12706314 in C7orf58 , P condition = 7 . 3×10−5 ) appeared , but did not achieve genome-wide significance . Using a similar conditional analysis ( with rs9525638 ) for the RANKL locus , no additional SNPs with an independent signal appeared . Two SNPs ( rs9525638 , rs2707466 ) were selected for replication in the MrOS Sweden cohort . In the replication stage , SNP rs2707466 at the WNT16 locus was significantly associated with tibial cortical thickness ( −0 . 11 SD per C allele , P = 0 . 008 ) , whilst no strong evidence of association was seen for rs9525638 near the RANKL locus although the estimated effect was in the same direction as in the discovery meta-analysis ( Table 1 ) . Thus , rs2707466 was the only SNP that was significantly associated with cortical thickness in both the discovery and replication cohorts ( combined −0 . 11 SD per C allele , P = 1 . 5×10−10 ) . Therefore , further analysis of associations with bone traits was constrained to rs2707466 at the WNT16 locus . Associations between rs2707466 and cortical thickness were highly similar when performed according to sex ( Figure 2 ) . No evidence of a significant impact of age for the association between rs2707466 and cortical thickness was found ( ALSPAC ( young ) vs . GOOD , YFS and MrOS combined ( adult and older ) : −0 . 09 SD vs . −0 . 13 SD per C allele , P = 0 . 135 , for heterogeneity between the two groups ) . In the combined meta-analysis , rs2707466 was not associated with either cortical vBMD or periosteal circumference ( Table S2 ) . In the GOOD cohort , rs2707466 was associated with cortical bone thickness also at the radius ( −0 . 12 SD per C allele , P = 0 . 008 ) . The information of the five forearm BMD cohorts is presented in Table S3 . A quantile-quantile plot of the observed P values showed a clear deviation at the tail of the distribution from the null distribution ( Figure S4 ) . The meta-analysis revealed that 54 SNPs within the 7q31 locus had genome-wide significant associations ( P<4 . 6×10−8 ) with forearm BMD ( Table S4 and Figure S5 ) . The most significant SNP was at rs2536189 ( −0 . 16 SD per C allele , P = 8 . 5×10−16 ) . Two common amino acid substitutions at WNT16 , rs2908004 ( Gly>Arg ) ( −0 . 16 SD per G allele , P = 1 . 2×10−15 ) and rs2707466 ( Thr>Ile as described in cortical thickness study ) ( −0 . 14 SD per C allele , P = 2 . 3×10−12 ) also demonstrated genome-wide significance ( Table 2 ) . The highlighted locus at 7q31 locus included genome-wide significant SNPs at the WNT16 ( wingless-type MMTV integration site family , member 16 ) , FAM3C ( family with sequence similarity 3 , member C ) and C7orf58 ( chromosome 7 open reading frame 58 ) genes ( Figure 3 ) . To identify the possible secondary signals in this locus , we carried out a conditional analysis . When conditioning on rs2536189 , the most significant SNP at WNT16 for BMD , an additional signal ( rs1554634 in C7orf58 , P = 7 . 8×10−8 ) was highlighted at a genome-wide suggestive level of association ( Figure S6 ) . This SNP is in LD with rs12706314 in C7orf58 ( r2 = 0 . 35 and D' = 0 . 72 in HapMap CEU ) , which showed suggestive association with cortical thickness when conditioning on the top signal for cortical thickness . In order to investigate whether the variants showing association with forearm BMD also have an effect on the risk of forearm fracture , we selected 4 genome-wide significant SNPs from the BMD analysis for de novo genotyping in samples with forearm fracture and their controls ( Table S5 ) , including the two missense SNPs in WNT16 ( rs2707466 and rs2908004 ) , one from FAM3C ( rs7776725 ) and one from C7orf58 ( rs10274324 ) . In the meta-analysis for osteoporotic fracture , comprising 2 , 023 forearm fracture cases and 3 , 740 controls , from 3 cohorts , we identified the rs7776725 SNP in FAM3C as being genome-wide significant for forearm fracture , with each C allele increasing the odds of fracture by 1 . 33 ( 95% confidence interval [CI]: 1 . 20–1 . 46 , P-value = 7 . 3×10−9 ) ( Table 2 and Figure 4 ) . The two missense SNPs in WNT16 also demonstrated strong associations with risk of fracture ( rs2908004 , risk allele G , OR = 1 . 22 [95% CI: 1 . 12–1 . 33] , P-value = 4 . 9×10−6 and rs2707466 , risk allele C , OR = 1 . 22 [95% CI: 1 . 11–1 . 33] , P-value 7 . 2×10−6 ) . SNP rs10274324 from C7orf58 was not associated with fracture in this study ( P = 0 . 15 ) . These results were consistent across the three cohorts ( Table S6 ) . Mice with a gene deletion of Wnt16 ( Wnt16−/− ) appeared healthy with no discernible morphological or growth defects , and had normal body weight and femur length at 24 weeks . In microCT analyses of the femoral diaphysis , male Wnt16−/− mice had a trend suggestive of reduced cortical thickness ( −7% , P = 0 . 14 ) , and reduced cortical bone polar moment of inertia ( −16% , P<0 . 001 ) ( Table 3 ) ; female Wnt16−/− mice had substantially reduced cortical cross sectional area ( −36% , P<0 . 001 ) and cortical thickness ( −27% , P<0 . 001 ) and calculated bone strength ( polar moment of inertia , −55% ) ( Table 3 ) . Trabecular bone volume fraction ( bone volume/total volume ) , as measured by microCT of 5th lumbar vertebrae ( LV5 ) , was similar in wild-type and Wnt16−/− mice ( Table 3 ) . In three-point bending tests , measures of bone strength ( stiffness , maximal force to breakage and work to failure ) were decreased between 43–61% ( 6 . 5×10−13<P<5 . 9×10−4 ) in Wnt16−/− female mice at both femur and tibia ( Figure 5 ) . However , microCT parameters of the femoral shaft and LV5 in male wild type and Fam3c−/− mice did not reveal any consistent differences across the three targeting strategies ( Table S7 ) .
Forearm fractures are a common and costly condition . In two separate GWASs for forearm BMD and cortical bone thickness we have identified variants that are genome-wide significant for these traits and , importantly , for forearm fracture at the 7q31 locus . Further , we have provided functional data from mice demonstrating that Wnt16−/− mice have reduced cortical bone thickness and bone strength . These results are among the first to demonstrate a genome-wide significant locus for osteoporotic fracture suggesting that this locus is an important genomic determinant of cortical bone thickness and forearm BMD and fracture as well . WNT16 is a member of the wingless-type MMTV integration site family , which has been reported to mediate signaling via canonical or non-canonical Wnt pathways . The canonical Wnt pathway has been shown to regulate bone mass . Specifically , loss of function mutations in the Wnt-co receptor LRP5 , as seen in osteoporosis pseudoglioma syndrome , result in a dramatic loss in bone mass [26] , while gain of function mutations give rise to extremely high BMD ( 5 SD above normal ) [27] . Wnt16 has been proposed to signal via the non-canonical pathway [28] , regulating haematopoetic stem cell specification in zebra fish , but whether this signaling system involves regulation of osteoblasts , which are of mesenchymal origin , is unclear . Little is known of the role of WNT16 in skeletal development and function , but Wnt16 has previously been implicated in synovial joint development in mice [29] . Several genes involved in the Wnt pathway have been previously identified to be associated with BMD by GWAS . These include known loci CTNNB [10] , SOST [30] , LRP4 [9] , [10] , LRP5 [8] , [10] , FOXC2 [10] , GPR177 [10] , and MEF2C [10] . Our study adds WNT16 to this list of bone-influencing Wnt factors . In addition to our GWA meta-analyses results for cortical bone thickness and forearm BMD , we present a functional study demonstrating that Wnt16−/− mice have a substantial decrease in cortical bone thickness ( 27% ) and bone strength ( 43–61% ) , but not bone length . Further , Medina et al ( accompanying submission ) provide data indicating that variation in WNT16 also influences BMD in children , suggesting that WNT16 may influence peak bone mass . Importantly , the clinical relevance of these findings at WNT16 is supported by our observation that WNT16 influences clinical fractures in humans and bone strength in mice . In an experimental study using forearms from cadavers , cortical bone thickness was highly correlated ( r = 0 . 93 ) to the 3-point bending failure load and could improve the prediction of this strength measure , in combination with bone mineral content derived from DXA [31] . Among individuals suffering a fracture at the radius , cortical bone thickness at the same site was 33% lower than in controls , which was the largest difference observed for the cortical bone traits [32] . The present study constitutes the first GWA study of cortical bone thickness , a trait with crucial importance for bone strength . The SNP rs2707466 , which causes a missense amino acid substitution ( Thr>Ile ) at the WNT16 locus , was consistently associated with cortical thickness in a meta-analysis of three large discovery cohorts and in a replication cohort ( combined P = 1 . 5×10−10 ) . In the GOOD cohort , the rs2707466 was also associated with cortical bone thickness at the radius , with an effect size similar to what was seen for the tibia , indicating that the WNT16 locus affects cortical bone thickness at both the forearm and leg . Interestingly , the most significant 14 of the 54 genome-wide significant SNPs in the forearm BMD GWAS , were all in the WNT16 or FAM3C genes , and showed high correlation with the two WNT16 missense SNPs: rs2908004 and rs2707466 ( 0 . 58<r2<1 , in HapMap CEU data ) . This finding might suggest that the association signals are driven by these coding SNPs in WNT16 . These two missense variants were also strongly associated with risk of fracture in our study , but did not achieve genome-wide significance . All together , our results implicate the WNT16 locus as important for fracture risk , which would likely be mediated via an effect on cortical bone , particularly the thickness of the cortical shell . FAM3C , which is predicted to be expressed in osteoblasts and encodes a newly identified cytokine necessary for epithelial to mesenchymal transition and retinal laminar formation in vertebrates [33] . We identified the SNP rs7776725 within the first intron of FAM3C to be genome-wide significant for forearm BMD ( P = 8 . 5×10−15 ) and forearm fracture ( P = 8 . 6×10−9 ) . Since the fracture cohorts do not have available BMD data , except for the AOGC cohort , which comprised only 7% of the fracture case population , no meaningful conclusions could be drawn for the independence of the association between fracture and forearm BMD . While candidate gene studies have previously described relationships between genetic variants and fracture [34] , [35] , [36] , we are aware of only one other variant that has been demonstrated to be genome-wide significant for any type of osteoporotic fracture , arising from the ALDH7A1 gene [12] . Interestingly , SNP rs7776725 in FAM3C was previously reported to be associated with speed of sound ( SOS ) as analyzed by quantitative ultrasound at the radius ( P = 1 . 0×10−11 ) in an un-replicated GWAS carried out in Asian populations [13] . This SNP was also associated with BMD in a Caucasian population [37] . The high-throughput DEXA and microCT screen which initially identified reduced cortical bone thickness and bone strength in Wnt16 knockout mice failed to observe any skeletal phenotype changes in three independent knockouts of mouse Fam3c . Since the sample size of Fam3c−/− mice was small ( N = 18 ) , the possibility of a false negative result cannot be excluded . All together , our functional studies indicate that Wnt16 rather than Fam3c is responsible for the observed genetic signal arising from this locus . However , we provide no data as to whether or not gain of functions variants in Fam3c could have effects on the studied bone traits and fracture risk . C7orf58 ( FLJ21986 ) , which codes for a hypothetical protein , has recently been identified to be associated with blood pressure in a study on Nigerians [38] . As an open reading frame , C7orf58 has no known function . In the forearm GWAS , the other 40 out of the 54 genome wide significant SNPs are from C7orf58 , and show low LD with rs2908004 and rs2707466 ( r2<0 . 2 , in HapMap CEU data ) , when conditioning for the top SNP ( rs2536189 ) in WNT16 , resulted in an additional signal ( rs1554634 ) in C7orf58 . Similarly , conditioning for rs2707466 at the WNT16 locus , in the GWAS for cortical bone thickness , resulted in an additional , suggestive signal ( rs12706314 , which was in LD with rs1554634 ) located in C7orf58 . Furthermore , Medina et al ( accompanying submission ) demonstrate in a conditional analysis that a separate signal , other than the signal derived from WNT16 , , located in C7orf58 was associated with total body BMD . Thus , these studies reveal an independent genetic signal for several bone traits , arising from C7orf58 , indicating a possible functional role of this protein . Even though our functional studies imply that Wnt16 determines the bone effects of the 7q31 locus , further studies are necessary to elucidate the role of C7orf58 . In summary , we provide the first evidence of association of common variants across the genome with cortical bone thickness , forearm BMD and forearm fracture . We also provide functional data implicating WNT16 at this locus . Importantly , our findings report one of two genome-wide significant variants for osteoporotic fracture . These results suggest a critical role of Wnt signaling pathway on cortical bone thickness and bone strength determination as well as fracture susceptibility .
All study participants provided informed written consent . Approval by local institutional review boards was obtained in all studies . 5 , 672 samples from five cohorts of European descent participated in this meta-analysis ( Tables S3 and S9 ) . BMD at forearm in all cohorts was measured by dual-energy X-ray absorptiometry following standard manufacturer protocols . Four genome-wide significant SNPs for forearm BMD were selected to test the association with forearm fracture in 2 , 142 cases and 3 , 697 controls from three cohorts . Forearm fracture was defined as fractures resulting from low trauma ( such as a fall from standing height ) occurring at the wrist , ulna , radius , forearm , as well as Colles' fractures . | Bone traits are highly dependent on genetic factors . To date , numerous genetic loci for bone mineral density ( BMD ) and only one locus for osteoporotic fracture have been previously identified to be genome-wide significant . Cortical bone has been reported to be an important determinant of bone strength; so far , no genome-wide association studies ( GWAS ) have been performed for cortical bone thickness ( CBT ) of the tibial and radial diaphysis or BMD at forearm , a skeletal site rich in cortical bone . Therefore , we performed two separated meta-analyses of GWAS for cortical thickness of the tibia in 3 independent cohorts of 5 , 878 men and women , and for forearm BMD in 5 cohorts of 5 , 672 individuals . We identified the 7q31 locus , which contains WNT16 , to be associated with CBT and BMD . Four SNPs from this locus were then tested in 2 , 023 osteoporotic fracture cases and 3 , 740 controls . One of these SNPs was genome-wide significant , and two were genome-wide suggestive , for forearm fracture . Generating a mouse with targeted disruption of Wnt16 , we also demonstrated that mice lacking this protein had substantially thinner bone cortices and reduced bone strength than their wild-type littermates . These findings highlight WNT16 as a clinically relevant member of the Wnt signaling pathway and increase our understanding of the etiology of osteoporosis-related phenotypes and fracture . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"endocrinology",
"genetics",
"and",
"genomics",
"diabetes",
"and",
"endocrinology",
"metabolic",
"disorders",
"clinical",
"genetics"
] | 2012 | WNT16 Influences Bone Mineral Density, Cortical Bone Thickness, Bone Strength, and Osteoporotic Fracture Risk |
The interaction of proteins at cellular interfaces is critical for many biological processes , from intercellular signaling to cell adhesion . For example , the selectin family of adhesion receptors plays a critical role in trafficking during inflammation and immunosurveillance . Quantitative measurements of binding rates between surface-constrained proteins elicit insight into how molecular structural details and post-translational modifications contribute to function . However , nano-scale transport effects can obfuscate measurements in experimental assays . We constructed a biophysical simulation of the motion of a rigid microsphere coated with biomolecular adhesion receptors in shearing flow undergoing thermal motion . The simulation enabled in silico investigation of the effects of kinetic force dependence , molecular deformation , grouping adhesion receptors into clusters , surface-constrained bond formation , and nano-scale vertical transport on outputs that directly map to observable motions . Simulations recreated the jerky , discrete stop-and-go motions observed in P-selectin/PSGL-1 microbead assays with physiologic ligand densities . Motion statistics tied detailed simulated motion data to experimentally reported quantities . New deductions about biomolecular function for P-selectin/PSGL-1 interactions were made . Distributing adhesive forces among P-selectin/PSGL-1 molecules closely grouped in clusters was necessary to achieve bond lifetimes observed in microbead assays . Initial , capturing bond formation effectively occurred across the entire molecular contour length . However , subsequent rebinding events were enhanced by the reduced separation distance following the initial capture . The result demonstrates that vertical transport can contribute to an enhancement in the apparent bond formation rate . A detailed analysis of in silico motions prompted the proposition of wobble autocorrelation as an indicator of two-dimensional function . Insight into two-dimensional bond formation gained from flow cell assays might therefore be important to understand processes involving extended cellular interactions , such as immunological synapse formation . A biologically informative in silico system was created with minimal , high-confidence inputs . Incorporating random effects in surface separation through thermal motion enabled new deductions of the effects of surface-constrained biomolecular function . Important molecular information is embedded in the patterns and statistics of motion .
Cell-cell interactions are critical in a variety of biological processes such as morphogenesis , immune responses , and homing . The interaction of surface-tethered biomolecules between cells is essentially two-dimensional because of the limited ability of the molecules to move in the dimension perpendicular to the cell surfaces . Receptors and ligands must therefore find each other by lateral motion on their respective surfaces [1] . A reactive head group attached to a macromolecular stalk extending from the surface of a cell results in a configuration with more factors governing function and more effects on cellular behavior than with the three-dimensional , freely-diffusive case . For example , in vascular homing the force response of the surface-tethered molecules is critical [2] . In cytotoxic T-cell mediated apoptosis , the two-dimensional , sorted arrangement of interacting partners might be important to developing a long-lived , death-mediating signaling complex [3] . Striking behaviors result from the complexity of surface-tethered molecules . The existence of catch-slip bonds has recently been demonstrated [4]–[6] . Catch-slip bonds exhibit the unexpected property that , as the force transmitted through the binding pocket increases , bond lifetimes increase prior to reaching a peak and then decrease . It has also been demonstrated that tether-constrained molecules more efficiently form bonds in some presentation contexts [7] . A two-dimensional molecular interaction system that has been studied is the binding of T-cells to antigen-presenting cells . There are many proteins involved in the intercellular interaction , and the binding of CD2 with LFA3 has been studied . As the membranes between the two cells remain in contact , they smooth against each other and form a space with a small separation [8] . Constraining the most likely position of the reactive sites to overlap well within the volume between the cells , or an even smaller space within the volume between the cells [9] , was found to result in up to a 40-fold enhancement in the effective reaction rate [8] . The molecular characteristics , such as length , flexibility of the molecular tether , and the binding pocket chemistry , that facilitate bond formation for two-dimensional interactions when the contact between surfaces is less than one second may be very different from the molecular characteristics that facilitate adhesion when the contact lasts minutes to hours . One example is cells traveling through the blood that capture to a blood vessel surface as a first step in homing to tissue . Vascular homing processes occur as white blood cells are recruited to sites of inflammation , lymphocytes travel to the lymph nodes , cancer cells metastasize to spread to new tissues , and stem cells home to sites of injury to repair tissues [10]–[12] . A cell traveling in excess of hundreds of cell diameters a second may briefly bump into the wall , leaving no opportunity for the proteins in the membrane of the flowing cell and in the membrane of the immobilized cell on the vessel wall to adapt for an optimal , sorted presentation of molecules . Although the average density of CD2 on T-cells is around 200 molecules per square micrometer [13] , within a factor of two of the average density of ligands mediating capture and dynamic adhesion on neutrophils [14] , the adhesive contacts involving hundreds of molecules per square micrometer have been observed to require thirty minutes to fully form in vitro with assays for CD2 and LFA3 [7] . T-cell interactions with antigen presenting cells have been observed to go through several phases in vivo , involving contacts lasting a few minutes and contacts lasting hours [15] , suggesting molecular sorting in the contact region plays a role in vivo . On the other hand , dynamic adhesion ligands are thought to be localized cellular membrane ruffles called microvilli [16] . The contact widths and times for these ridges during cell capture are much shorter and smaller than for CD2 and LFA3 , as small as 100 nm and as short as 1 ms [16] , [17] , respectively . Also , once they form , the reacting pairs must survive higher forces exerted on the cell . Consequently , there may be a specialized set of structural , dynamic , and kinetic features of the molecules responsible for cell capture that facilitate rapid molecular tether formation and lifetime . Selectins mediate dynamic interactions between cellular surfaces . Selectins have received considerable attention because of their importance in inflammatory and immune trafficking as well as their role in diseases such as atherosclerosis and cancer metastasis [10] , [18] . Many assays have been employed to make measurements of selectin molecular interactions: laser traps , atomic force microscopy , biomembrane force probes , and flow cells [5] , [19]–[23] . Arguably , a significant advantage of flow cells is that they give a report of molecular binding that is quite functionally relevant . Flow cell assays capture the characteristics of a hydrodynamic environment more directly than single-molecule assays . They balance the experimental complexity of an in vivo vascular model and the ability to make deductions about biomolecular interactions at the most basic level . Observations of complicated cellular behavior in a flow cell , such as hydrodynamic shear thresholding , have helped to inspire the application of force spectroscopy techniques that have established the existence of catch-slip bonds [24] , [25] . It is not clear which known qualitative molecular characteristics are important to their functional ability to capture a cell or particle and initiate bonds that can withstand detaching forces . We therefore adapted an adhesive dynamics modeling strategy that can test functionally relevant P-selectin/PSGL-1 molecular behaviors . Novel aspects of the simulation and analysis methodology were: The simulation results and analysis methodology resulted in several new findings: There were also two findings where qualitative changes in the expression governing molecular behavior did not make a difference to the simulated sphere's motion . The effects of molecular confinement on bond formation were not functionally important to dynamic adhesion . The result demonstrates a criterion where successful static and dynamic interaction systems differ . Also , P-selectin/PSGL-1 catch-slip bonds performed nearly as well as slip-only bonds in mediating capture interactions , and were equally effective at mediating pauses . The result reinforces the hypothesis that a purpose for catch bonds might lie in distinguishing soluble ligands from immobilized ones rather than regulating the dynamics of adhesion , at least for P-selectin/PSGL-1 . The implications of the present study are extensive . With an increased understanding of the molecular features that enhance bond formation between surfaces , it will be possible to engineer molecular systems with an optimal physiologic impact . Optimizable systems include enhanced targeted drug delivery and molecular imaging agents and dendritic cell therapies with the potential for enhanced T-cell activation .
For each simulation , discrete attachment points for the base of individual adhesion receptor molecules were randomly distributed over the surface of a microsphere . To accurately capture scenarios where the adhesion receptor was immobilized to the surface , as with experimental microbeads or proteins anchored to the cytoskeleton , on-sphere diffusion of the anchorage points was assumed to be zero . A stochastic methodology was employed to include three-dimensional lateral , vertical , and rotational diffusion of the sphere and is described in greater detail in Protocol S1 . Three-dimensional diffusive motion of the molecular binding pockets within the contact volume was treated during consideration of the on-rate expression . The simulation geometry is illustrated in Figure 1 , and steps in the calculation are detailed in Figure 2 . Two models were compared to account for different mobilities of the molecular binding pockets on their respective tethers . In the first , it was assumed each receptor within a bond length of the surface could find a ligand with an equal rate: ( 1 ) The symbols and values employed are defined in Table 1 . The Heaviside function , denoted by H ( ) , limited nonzero bond formation rates to receptors in the contact patch . The reaction rate described by ( 1 ) could be described as contact patch confinement because only receptors in the contact patch could react , and finer details of the reaction configuration were assumed to be unimportant for receptor function . The concept is illustrated in Figure 1B . In the second approach , receptors with anchorage points on the sphere closer to the surface were assumed to sample a wider area of the surface , enhancing their probability of encountering an immobilized binding partner: ( 2 ) We refer to the reaction model described by ( 2 ) as molecular area confinement . The geometry of the search by the receptor's binding pocket for a ligand immobilized within a suitable distance influences the reaction . Therefore , ( 2 ) is conceptually similar to the geometric interpretation of the enhanced apparent association rate of laterally diffusive cellular CD2 with LFA3 on two surfaces with an extended contact time [7] . According to ( 2 ) , reactions proceed more quickly when the receptor is held closer to the surface and sweeps out a broader area in the search for ligands . The concept is illustrated in Figure 1C . The rate , kfT , was normalized based on kf so both ( 1 ) and ( 2 ) would yield the same average reaction rate across the contact patch for a sphere touching the wall to within surface roughness limitations . A number of models of receptor-ligand dissociation kinetics under an applied load have been developed to describe the dissociation of selectins from their ligands . One of the first models , based on observations of non-covalent solid materials failure [26] , proposed that the application of force increases the dissociation rate in an exponential manner [1] . Bonds exhibiting this type of force response kinetics are referred to as “Bell slip bonds” to distinguish the model from alternate proposals for the forced dissociation relation [27]–[29] . Increasing force causes the slip bond to dissociate more quickly . Newer measurement techniques demonstrated force decreases selectin bond dissociation rates until a peak mean lifetime is reached , and higher levels of force increase the bond dissociation rate . Bonds exhibiting this type force response kinetics are referred to as “catch-slip” bonds because of the biphasic force response . Several theoretical and mathematical models have been developed to describe catch-slip bonds [22] , [25] , [30] , [31] . The five-parameter model of rapid internal state equilibration [22] is an appealing model that has a sound theoretical basis , captures the salient quantitative features of selectin dissociation kinetics [32] , and converges on a high-impedance dissociation pathway at high forces that closely resembles the Bell slip bond model of dissociation . The dissociation rate for existing receptor-ligand complexes was first computed using the Bell slip bond model for dissociation under an applied load [1] , [26]: ( 3 ) After initial validation and investigation with slip bonds , the effect of catch-slip bonds on motion was investigated utilizing the five-parameter model of rapid internal state equilibration [22]: ( 4 ) It was assumed a bond exerted no force when in compression but behaved as a very stiff spring when stretched past its contour length . A constant dissociation rate for a bond in compression has been assumed in previous simulations of selectin-mediated rolling [33] . We refer to the model combining this assumption with a stiff spring for extension [34] , [35] as a rope model . The expression employed was: ( 5 ) Note that the rope model spring constant was very high , so bonds did not extend much past their equilibrium length . Alternatively , a freely-jointed chain model was used with the condition that a chain in compression exerted no force: ( 6 ) Although the time steps were much smaller , data from the model was sampled to file at 1 , 000 fps , in analogy with experiment . This represents an upper sampling limit for many experimental systems used to acquire data optically . Numerical parameter values used in the simulation are given in Table 1 . The force deformation models are described in greater detail in Figure S1 .
Sphere motion in the absence of binding interactions established a baseline for both validation against experimental results and comparison to reactive sphere motion . Vertical stepping accuracy was first investigated by recording the gap and velocity distribution of vertically diffusing microspheres over a long time interval in the presence of gravity and fluid flow . Figure 3 compares the results for motion between a 6 µm-diameter and a 10 µm-diameter sphere in the presence of a 50 s−1 wall shear rate for 1 , 000 s of simulated time . The results for vertical motion are shown in Figure 3A . As in all simulations presented here , a minimum separation distance of 10 nm was enforced with a reflective boundary condition to simulate a minimal wall roughness that would be experimentally achievable and avoid singularities in the fluid dynamic equations . The data for the simulated microspheres agreed very well with the theoretical Boltzmann distribution for particles diffusing in a gravitational force field , validating the vertical stepping method . Thermal fluctuations in the vertical coordinate were more significant for smaller particles than for larger ones . The length scales for vertical motion in Figure 3A are biochemically relevant because biomolecular interactions occur on the nano-scale . Of special interest is the PSGL-1/P-selectin bond , which has been estimated to extend up to roughly 90 nm when unstressed [36]–[38] . The simulation results predict the 10 µm-diameter and 6 µm-diameter spheres will remain with 90 nm of the surface 99% and 67% of the time , respectively , in agreement with the Boltzmann distribution . Only after 1 , 000 s of simulation time did the probability distribution function converge well upon the equilibrium distribution . The long time required to achieve statistical equilibrium suggested the importance of transport history for interacting spheres . The long equilibration time has implications for experimental flow cell studies employing microbeads . First , all of the beads that appear to be near the wall are not equivalent . Some beads spend a larger portion of their time near the surface than others while traversing a field of view . Some beads are sufficiently far so they are incapable of binding even if the population is given sufficient time for sedimentation . Secondly , a flow cell must be very long , 10 cm , for the beads in a field of view to follow the equilibrium distribution if it takes 1 , 000 s for a 10 µm-diameter bead to be able to fully sample the gap possibilities once the bead sediments to near the surface . The probability distributions for the sampled flow-direction velocity , VS , X , for the non-interacting spheres flowing at 50 s−1 are shown in Figure 3B . Variation in the convective component of the motion was expected due to variations in the vertical coordinate of the sphere . It is stressed that these are sampled velocities: there was a convective as well as an effective diffusive component to motion in the flow direction . As the gap between the bottom of the sphere and the wall increased , the sphere's velocity increased due to the linear shear gradient . The apparent flow-plane diffusive component of the sampled velocity may be positive or negative . The standard deviation of the diffusive component of the sampled velocity will decrease as the sampling rate is decreased ( refer to Protocol S1 ) . As might be expected , the mode of the sampled velocity data for the 6 µm-diameter sphere was lower than the mode for the 10 µm-diameter sphere due to the smaller size . The sampled velocity for non-interacting , 6 µm-diameter spheres flowing at 50 s−1 was skewed from normal , very similar to the pooled experimental instantaneous velocity results as previously reported [39] . The mode of the simulation results , 90 µm/s , was higher than the experimental mode previously reported [39] , 70 µm/s . The discrepancy was still within a range that can be accounted for by experimental differences in the microbead size , possible variations in the observed populations , and differences in the surface roughness and coatings . The sampled velocity results for the 10 µm-diameter sphere were less skewed because there was a smaller variation in the distribution of gap size for the larger sphere . A statistical plot of VS , X against the gap size is presented in Figure 3C for the 6 µm-diameter and 10 µm-diameter spheres . During each time step , the forces , torques , and damping factors acting on the sphere were computed by interpolating the individual fluid dynamic solutions and then applied using superposition . The deterministic superposition solutions previously tabulated [40] are shown on the same plot . There was a slight disagreement between the statistical mode for the simulation and the tabulation . Observing vertical slices through the “Gap Size” axis , it is apparent that the mode of VS , X as a function of gap size is less than the tabulated value reported by Goldman et al . [40] . There are two possible reasons for the discrepancy . The first may be a cumulative effect of residual errors in the interpolation method used to obtain the individual hydrodynamic damping factors . The other is that the statistical weighting of motions truly results in a mode that is lower than the deterministic solution . Such differences have been theorized to occur with important consequences in biomolecular reaction systems [41] . Despite the minor discrepancy , the results agreed sufficiently for present purposes . Figure 3D presents sample time domain VS , X data for the both sphere sizes . The vertical excursions of the 10 µm-diameter sphere away from the wall were infrequent and of small magnitude . The high-frequency fluctuation in VS , X was largely due to the high sampling rate: lateral diffusion was well distributed across frequencies . The low-frequency component of the velocity fluctuations agreed well with variations in the gap size . The effects of microsphere size and the molecular force distension model were compared to gain a quantitative understanding of how contact area and molecular distension affect motion . Figures 4A–C present simulation results for spheres bearing 50 sites/µm2 of receptor interacting with 100 sites/µm2 of ligand at a wall shear rate of 50 s−1 . The data for Figure 4 was sampled at 1 , 000 fps . Figure 4A presents simulation results for a 6 µm-diameter sphere that forms rope-like bonds , Figure 4B presents data for a 6 µm-diameter sphere that forms freely-jointed chain bonds , and Figure 4C presents data for a 10 µm-diameter sphere that forms freely-jointed chain bonds . As an additional tool to validate and interpret the physics , three-dimensional videos were constructed from the simulation results ( sampling reduced to 250 fps , Videos S1 , S2 , S3 ) . One second of data was selected from each of these scenarios when constructing the videos . The videos illustrate the microsphere behaved in a physically realistic way . Bond formation events introduced realistic forces and torques that caused the sphere's rotational orientation and centroid to converge on a stable mechanical equilibrium point . The sphere was still free to undergo stochastic fluctuations in translation and rotational orientation when settled in the bound state . Roughness held the sphere 10 nm from the absolute , mathematical surface , but did not prevent the sphere's oscillations . In practice , the motion of a bead being pushed into the surface by a biomolecular lever arm depends on the details of the experimental surface construction and blocking strategy . The results must be construed as a case representing an ideal experimental methodology . The results recreated the discrete “stop and go” behavior observed with microbeads and demonstrated the effect of differences in biomolecular tether properties . Discrete stops were apparent in Figures 4A–C as drops to a zero-mean , fluctuating VS , X . The results closely capture the discrete pause behavior reported for selectin-coated microbeads [39] . The results also illustrated differences in molecular tether stiffness can influence the sphere's motion during a binding event . It is important to point out the interpretation of experimental velocity fluctuations must also include a consideration of noise in the acquisition system [42] . Fluctuations in VS , X in for the sphere bound by the freely-joined chain in Figure 4B were larger than the sphere bound by the rope in Figure 4A , although fluctuations in VS , Y agreed more closely . Although there were significant fluctuations in the gap between the bottom of the sphere and surface prior to the first bond formation event , the sphere maintained contact following the first capture event . Multiple bonds were often present , but most frequently only one bond supported the sphere against the hydrodynamic load . The bond loading forces from the trials in Figures 4A , B are presented in Figure 4D . The top panel illustrates the result with the rope model and the bottom panel illustrates the result with the freely-jointed chain model . Insets show the results for the second bond formed in each trial and better illustrate the force fluctuations by expanding the time axis . The magnitude of the force fluctuations with the stiffer rope was apparently larger than for the freely-jointed chain , but the average force on the bond was similar . The theoretical result was interesting because , if such fluctuations occur , they would result in transient peak forces larger than the mean calculated from the average bond angle and shear force . Forces experienced by the bonding pocket have an effect on molecular conformation and function [25] . The result suggested tether properties can be transmitted to the binding pocket and may influence function . Also , many bonds formed and dissociated without ever supporting a hydrodynamic load , in agreement with simulations of leukocyte rolling [43] . Although less common , several multivalent force-bearing bond events occurred . For the trial in the bottom panel in Figure 4D , this is apparent shortly before 5 s: simultaneous load bearing bonds supported a lower peak force . Video S3 demonstrates how the sphere can arrive at such a configuration: multiple force-bearing bonds also occurred with the 10 µm-diameter sphere at 7 . 920 s . Bonds near the edge of the contact patch can also become stressed due to stochastic excursion in the sphere's position: brief loads are apparent in the bottom panel of Figure 4D , such as at 8 . 972 s . Video S2 illustrates how this is possible . Bonds shorter than the deduced unstrained bond length supported no force and could form freely . While the sphere paused , it was possible for additional bonds to form in the contact patch . Although the bonds must form while unstressed , the sphere could still undergo small diffusive fluctuations in position to stress them . Mechanical loading history can be an important factor governing bond lifetimes [44] . The molecular force loading history was explicitly investigated . A simplifying assumption was made in the model: bonds shorter than the deduced unstrained bond length supported no force . As the bonds extended further , there was a step in the force/length relationship . The assumption amounted to a step to 56 pN as the length extended past the unstressed molecular contour length of 92 nm , as calculated from freely-jointed chain model parameters [21] , and increased continuously thereafter . The force/length relationship is presented in detail in Figure S1 . Bond loading results from different simulations are presented in Figures 5A–C . Bonds were temporally aligned so loading would occur at 0 . 001 s . The larger , 10 µm-diameter spheres had a larger area within a molecular contour length of the surface , tended to form the most bonds , and exhibited the most frequent multiple bond formation events . Figures 5A , B illustrate the result for a 6 µm-diameter sphere interacting with a 50 s−1 and 100 s−1 wall shear rate , respectively . Although the peak load was about double in the latter case , the increased shear did not have as much of an effect on the loading rate . Also , there was a multivalent bonding event apparent with the blue and green tracings , which explains why they were not individually loaded to the full force required to restrain the sphere against hydrodynamic drag . Figure 5C illustrates the result for a 10 µm-diameter sphere interacting with a 100 s−1 wall shear rate , which also has a much larger contact area . Note that aside from many more binding events and more multiple loading events , some of the force tracings exhibit a prominent inflection point between 0 . 001 s and 0 . 004 s . This feature is also present , although less pronounced , in Figure 5B . Figure 5D demonstrates bonds only become loaded and exert lateral motion on the sphere towards a mechanical equilibrium point once they leave the contact patch . Results from 10 µm-diameter spheres interacting with a 100 s−1 wall shear rate were screened for singly-loaded bonds and the force was plotted in relation to the receptor attachment point relative to the sphere's center . The result explains the inflection-like point in Figure 5C sometimes apparent for up to the first 5 ms of bond loading . Single bonds can become stretched just until the point they exert a restoring force , 56 pN , in the contact patch as their anchor point on the sphere rotates into the upstream hemisphere . Bond loading increases rapidly once the anchor point on the sphere rotates out of the contact patch and can then induce motion perpendicular to the flow direction , wobble . Figure 6 demonstrates increasing shear has a small effect on mechanical equilibrium during a pause event , where molecular bonds restrain motion of the sphere against hydrodynamic forces . The time to achieve the equilibrium configuration is slightly increased with increased shear . Singly-loaded binding events were pooled from simulation runs for 6 µm-diameter spheres with wall shear rate of 50 s−1 , 6 µm-diameter spheres at 100 s−1 , 10 µm-diameter spheres at 50 s−1 , and 10 µm-diameter spheres at 100 s−1 . As expected , the mean peak loading force increased with wall shear rate and was 88 , 156 , 247 , and 428 pN , respectively . Note that these values correspond to the maximum of the force fluctuations for single bonds as observed in Figure 4D . It is worth noting these forces would cause dissociation of the adhesion molecule from the cytoskeleton in leukocytes and therefore decrease force on the bond [45] , [46] . The peak bond force did not exhibit a strictly linear dependence on shear . The freely-jointed chain tether distensibility allowed the tethers supporting the bond to extend slightly and the lever arm angle to decrease as increased shear force increased the biomolecular distension . In all conditions , the standard deviation of the peak bond force was small: less than 4 pN . Increasing the wall shear rate had a small effect on the mean peak single bond loading rates . They were 61 , 71 , 62 , and 96 pN/ms , respectively . Having demonstrated the simulation qualitatively recreated motions observed for non-interacting spheres and also recreated the discrete , transient stops observed for interacting spheres , the next important consideration was whether the results matched detailed motion patterns observed experimentally . The site density , measured bond force-response characteristics , sampling rate , and diameter were chosen to match the previous investigation [23] . Given the alternative bond lifetime models and parameter discrepancies in the literature , several alternative dissociation models and rates were selected to test whether they might give a match to experiment . A brief comparison of alternative models of bond dissociation for P-selectin/PSGL-1 is shown in Figure S2 . Two methods were employed to judge the quality of the match between the simulation and experiment . In the first , a video sample was obtained from Dr . Eric Y . H . Park . Simulation results were screened by eye to identify a period with similar qualitative behavior to a small experimental tracking data set . A detailed comparison of the motion was made . Secondly , velocity tracings from simulated microbeads were analyzed using a pause time analysis method to deduce an experimentally apparent , effective dissociation constant . Complete velocity results from the simulations employing parameters derived from the experimental study [23] are presented in Figure S3 . Differences between the experimental and simulated motion patterns that suggested a refined interpretation of the experimental data needed to be made . Figures 7A , B present experimental data from the same microbead tracked by the two different methods . Figure 7A presents experimental tracking data from the previously published analysis [23] using a sum-of-absolute-differences algorithm . Figure 7B presents experimental results using a centroid tracking algorithm with intensity threshold segmentation , MCShape . The additional comparison set facilitated interpretation of the experimental results by establishing confidence that the velocity waveform characteristics were not noise artifacts resulting from the tracking method . The blue tracings show VS , X and the green tracings show VS , Y . Figure 7C presents experimental data from an apparently non-interacting particle for reference purposes . Figure 7C illustrates the magnitude of the noise that can be expected from the algorithm employed in Figure 7B . Although some noise may be present in the velocity signal in Figure 7B , there is clearly a long pause beginning at roughly 0 . 35 s . Additional , briefer pauses are apparent with both tracking methods in Figures 7A , B . Figure 7D illustrates a selected portion of simulation results using the Bell dissociation model parameters reported for microbeads [23] , sampled at 250 fps . The reduction in simulation sampling rate from 1 , 000 fps to 250 fps reduced the fluctuations in the sampled velocity due to diffusion . The simulation qualitatively recreated the starting and stopping events observed in the experiment , although there were fewer very short pauses in the simulation results than observed experimentally . The deceleration to a near-zero VS , X agreed with experiment well , and the particle took several frames to slow in both cases . The simulation missed the lagging component in the acceleration that was apparent in the experimental data . A detectable lag period required for particle acceleration was observed previously with detachment from ligand-presenting accumulation strips [47] . There were several brief deceleration events in the experimental tracking results in Figures 7A , B that were larger than the noise in Figure 7C but missing from the simulation results in Figure 7D . The disagreement suggested an important component to the experimental physics was missed in the analysis and therefore not included in the simulation . Pause time statistics calculated from simulated sphere motion matched experimental results . The statistical point estimate koff , an indicator of the dissociation rate for individual molecules or molecular clusters loaded with force , was calculated as described previously [48] . Inputting the Bell model molecular parameters experimentally measured for microbeads into the simulation [23] , an apparent koff was obtained from the simulated velocities that matched the statistical point estimate to within 8% ( Table 2 ) . There have been many measurements of selectin kinetics and mechanical responses . The reported measurements vary by orders of magnitude . Analyses have incorporated receptor multivalency [22] , [48] , [49] as well as cellular deformability and microvillus elongation [23] , [48]–[51] as possible reasons for the discrepancies . The experimental result of Evans et al . [22] represents a monomeric bond formation case . Simulation results assuming parameters estimated from the discussion [22] are presented in Figure 7E . The result demonstrates monomeric bonds would not result in pauses , at least for 10 µm-diameter spheres . Some bond events might not even be detectable above noise . Comprehensive results from the simulations employing the monomeric parameters [22] are presented in Figure S5 . The results suggest the transient deceleration events in Figures 7A , B that did not pause the sphere could be low-valency bond formation events . A conclusion of previous studies has been that dimerization plays a significant role in measured cellular bond lifetimes [48] , [49] . We refer to multivalent molecular groupings that form bonds as a unit and evenly distribute a force as clusters . Reliability theory rules governing cluster dissociation , similar to those employed in previous analyses of bond lifetime [52] , were added into the present model . The goal was to investigate whether bond clusters could account for the observed discrepancies in the flow cell microbead pause kinetics with the parameters measured by molecular force spectroscopy techniques . Clusters were assumed to form at the same rate as monomers . This assumption facilitated the interpretation of the motion statistics , although dimerization has been reported to result in a two-fold enhancement on bond formation rates , as assessed by detected pause events [53] . The summary statistics for a variety of simulation conditions are compiled in Table 2 . A more comprehensive compilation of results is available in Table S1 . Reliability theory was used to create a dimeric grouping of the catch-slip parameters obtained from experiments with dimeric P-selectin/PSGL-1 interactions [5] , [32] , which might physically correspond to tetrameric bond clusters . The statistical point estimate obtained for the dimeric dimers indicated dissociation kinetics still faster than observed experimentally . The observed koff calculated from the simulation also closely matched the statistical point estimate . Trimeric groupings of the catch-slip dimers [5] , which might correspond to hexameric clusters , produced dissociation kinetics slightly slower than experiment ( Table 2 ) . Simulation results suggested the flow cell experiment [23] was primarily detecting dimeric to trimeric groupings of dimers measured in the force spectroscopy experiment [5] . The cluster had to be increased to 3× dimers for the observed koff to approach that reported by Park et al . [23] ( Table 2 ) . Interestingly , membrane P-selectin has been reported to form noncovalent hexamers under some isolation conditions [37] . Membrane PSGL-1 has also been observed to form rosettes [36] . A thorough analysis optimizing cluster size distributions to match bond lifetime data has been previously performed for cellular systems [48] , [49] . Therefore , subsequent analysis investigated what might be expected from an experimental microbead flow system similar to the previous study [23] , except with the molecules immobilized in a dimeric configuration . Fluctuations in VS , Y , wobble , might also contain information about biomolecular tether formation events . The simulation results in Figure 7D demonstrate brief increases in the magnitude of VS , Y at the same time VS , X is observed to decrease . The wobble was not readily apparent in the experimental velocity results . The tracking results in Figure 7A exhibited little variation in VS , Y . The algorithm employed in Figure 7B exhibited random variations in VS , Y that appeared to be noise . There is one event just before 0 . 7 s that might correspond to a real wobble . The movies were taken with a 20× objective and the movie quality would be improved with current technology . It is possible better resolution will detect real wobble . Higher site densities do not mediate extended pauses nearly as well as when the receptors and ligands are packaged into molecular clusters , as shown in Figure 8 . The catch-slip parameters regressed from the dimeric flow cell data [5] , [32] were used for further analysis . They represented a minimum achievable valency configuration for experiments employing wild-type PSGL-1 and were able to detectably pause the sphere . Comprehensive velocity results with these parameters are shown in Figure S4 . The effect of receptor distribution was tested by increasing cluster valency or alternatively increasing the number of receptor clusters . The apparent koff was estimated from the slopes of the black lines in Figure 8C . Some quantization was apparent in the pause time values due to the brevity of the pause relative to the sampling rate . The receptor cluster site density basis was 95 sites/µm2 . The results with 1× and 2× receptor cluster density were similar for an assumed cluster valence of one , suggesting multiple hydrodynamic load-bearing attachment points could not form efficiently . It was necessary to package receptors and ligands into molecular clusters to effectively extend pause times . Recent observations suggest that skip distances are an important measure of biomolecular binding efficiency [39] . An analysis of how far the modeled sphere traveled between the pause events was performed . Single-component Poisson models could not reconcile the initial steep slope and shallower response phase at longer skip distances apparent in Figure 8B . A logarithmic transformation was employed , as shown in Figure 8D . Two linear segments were apparent in the transformed data , suggesting a statistical model blending multiple Poisson processes might match well . The mixed Poisson process model was tested: ( 7 ) where ( 8 ) Here , d is the skip distance , the Pi's are the probability of one of the N Poisson processes , and the Δi's are the respective rate parameters with dimensions of distance . Parameter estimates were derived using nonlinear regression in MATLAB . The results for a two-component Poisson process are plotted as black solid lines in Figures 8B , D . The regression fit the data well . A physical explanation for the high probability of short skip distances relative to a single-component model would be the existence of pre-existing bonds in the contact patch . When the hydrodynamic load-bearing bond breaks , the sphere could only perform a long skip if no pre-existing bonds were present in the contact patch to catch the sphere . Indeed , the short-skip distance derived from the regression was on the order of the size of the patch where molecules on the sphere could contact the surface , one micrometer ( Table 2 ) . Figures 8B , D illustrate , as expected , the skip distance was most effectively reduced by increasing the density of receptors on the surface . Surprisingly , doubling the valency with a constant cluster density was almost as effective at reducing the skip distance as doubling the site density with constant valency , despite a constant association rate . The result demonstrates dissociation kinetics can influence measures of bond formation . The result also reinforced the conclusion that functionally effective molecular interactions require clustering . To investigate the effects of the catch component in catch-slip bond formation , simulations were also run assuming the high-impedance pathway parameters derived from the dimeric flow cell study [5] , [32] , entered as a Bell slip model . The results are indicated as “slip only” in Table 2 . If catch-slip bonds were present , it was postulated they might result in a decreased sensitivity of the short-skip fraction to the amount of ligand available . A test of the functional impact of reaction enhancement due to confinement was included in the analysis by comparing the results with different receptor bond formation rules defined by ( 1 , 2 ) . Reactions enhanced by molecular area confinement might be expected to result in smaller short-skip distances than reactions exhibiting contact patch confinement . Instead , a distinguishable effect of the confinement model on the short-skip measurements was not observed . This may be due to the smaller sample size at lower ligand densities . However , a robust indicator of biomolecular activity should have a detectable pattern even with the smallest sample size of 24 events . Instead , an emerging trend in the long-skip data was apparent in Figure 9A . Surprisingly , the long skip distances clustered better according to the functional form of the off-rate than the on-rate at low ligand densities . At the lowest ligand density , a roughly 25% reduction in the long skip distance was observed with the “slip only” bonds . The results emphasize the difference between molecular tether formation and pausing . The differences due to the assumed off-rate model suggest that when interaction is mediated by a small number of bonds , the dissociation kinetics influence the ability to initiate a pause . A reasonable explanation for the observed decrease in the long-skip distance with the removal of the catch component is that the catch component increased the dissociation rate of transient molecular tethers before they could rotate out of the contact patch , become stressed , and effectively pause the microsphere . The difficulty in making deductions about molecular confinement from skip distances with the selected particle size and dissociation parameters suggested another experimental metric was necessary . Therefore , we investigated whether motion perpendicular to the flow direction , wobble , carried information regarding two-dimensional formation kinetics . The probability distribution function for VS , Y was identical at a given ligand concentration , regardless of the assumption of the functional form of the molecular formation or dissociation rate ( data not shown ) . The identical probability distribution functions for VS , Y suggested diffusive motion obscured the analysis . The persistence of a wobble was investigated using autocorrelation . It was reasoned that bond-directed rather than diffusive-directed motion should correlate over short time intervals as a bond became stressed . The autocorrelation of VS , Y yielded informative results and is presented in Figure 9B . At low ligand concentrations , the wobble autocorrelation grouped very well by the assumed confinement model . Confinement-sensitive biomolecules wobbled the sphere less , as their tether anchor points on the sphere were more likely to be more proximal to the center of the sphere's planar projection . As ligand concentration increased , the dissociation model also played a role , although smaller , in the particle wobble . The role of surface separation in initial capture and recapture events was investigated . The results from the simulations investigated in Figure 9 were pooled . A capture bond was defined as a bond that formed when there were no existing bonds in the previous time step . The distribution of gap sizes during the first capture bond and subsequent recapture bond events is presented in Figure 10A . The distribution was pooled from 60 simulations , and the result for the first capture bond agreed relatively well with the Boltzmann distribution governing the separation at equilibrium . Therefore , at the site densities employed , the PSGL-1/P-selectin pair effectively reached across the 90 nm gap to mediate initial bond formation . There was not a detectable requirement for the sphere to undergo a thermal excursion closer than the molecular contour length to form a bond . Bond recapture events , which would roughly correspond to “long skips” in Figure 9B , were observed to occur at smaller separation distances . The result suggested that subsequent bond formation events should occur more quickly than the initial because more receptors would be within an unstressed molecular contour length of the wall , as suggested by ( 1 , 2 ) . In Figure 10B , the time until the first capture event is plotted as well as the interval between bond breakage and recapture . Indeed , recapture occurred more quickly than initial capture . The lateral motion of the stationary and moving surface relative to each other , as observed in the rolling of leukocytes , can affect the rate of reaction in some situations and merits specific consideration . A departure from the previous analysis of lateral relative motion [54] would need to be implemented for the process investigated presently for two reasons . First , a small number of bonds with significant changes in their relative number can be observed in Figures 4A–C , demonstrating the present process is not near steady-state and is therefore inconsistent with the assumptions of the previous analytical model [54] . Secondly , the appropriate diffusivity model for molecular binding pockets firmly attached to an immobile anchor point on a surface by a tether is qualitatively different than one in which the tether attachment point is also free to diffuse in a membrane . To better elucidate these points , consider the general equation for convective and diffusive transport: ( 9 ) For initial biomolecular bond formation , or for the case where there are significant fluctuations in the free ligand density due to the stochastic nature of bond formation , the time derivative of the concentration will not be zero as assumed in the previous relative motion analysis [54] . The transition from no bonds to at least one bond , as during initial tethering , is not a steady-state process , and it will be desired to accurately capture this step . Secondly , it is not intuitively clear what the diffusivity constant in ( 9 ) represents if the reactive end groups are free to diffuse about a tether point but the surface attachment point is not mobile in a membrane . As the sphere rotates , receptors that have formed bonds or are engaged in encounter complexes must dissociate for the sphere to move forward because the receptor attachment points are fixed on the sphere's surface . A more detailed treatment is presented in Protocol S2 . Therefore , relative motion cannot enhance bond formation by lateral transport effects when the receptors are immobilized to a point on the surface of the sphere and ligands are also immobilized on a surface , as in the present case . It is possible that relative motion might decrease bond formation . If the sphere moves fast enough such that the receptor and ligand binding pocket , once they happen upon a suitable encounter , cannot complete reorientation of residues in the binding pocket to complete the bond , no bonds can form . The requirement for bond formation introduced by the consideration of relative motion is: ( 10 ) This portion of the motion analysis is similar to previous studies [54] , [55] . VXYslip , R is the slip velocity a receptor on the sphere relative to the surface , and VXYslip , R is less than the sphere's velocity due to the rotation of the sphere . The reaction rate , r+ , is similar to previously described intrinsic reaction rates [1] , [54] , [55] , except an additional transport step can be removed and made explicit , as discussed in Protocol S2 and shown in Figure S6 . The intrinsic bond formation rate should be very fast , and the timescale has been projected to be around 10 µs or less from simulations [54] . The quantity on the left in ( 10 ) should be around 1 , 000 µs for the shear rates employed here . A decrease in reaction due to relative motion would not be expected for the shear rates employed .
Several key findings were made in the present investigation . The first two of these were especially apparent through a detailed analysis of sphere's motions in the simulation . First , the grouping of molecules into load-sharing clusters is critical for function . Single bond formation events cannot pause the sphere at the wall shear rates investigated because a single receptor-ligand bond cannot withstand the force . Secondly , the wobble autocorrelation may serve as an indicator of confinement enhancement in the molecular formation kinetics . Finally , it was also observed experimentally measured P-selectin kinetics and densities are able to effectively capture a particle as long as the particle is within a molecular contour length of the surface . Furthermore , recapture is enhanced by the proximity to the wall . The simulation method presented here differs in several important fundamental ways from previous computational models of adhesive interactions in flow . A theoretical framework for modeling the vertical and lateral diffusion of microspheres under flow was previously developed [56] , but the previous investigation did not incorporate biomolecular bond formation . The implementation presented here also adds rotational motion and rotational diffusion , since they were needed to track the position of individual receptors and molecular tether attachment points . Previous work developing adhesive dynamics simulations provided an invaluable reference and a presentation of many of the components of the physics employed [57] . The inclusion of thermal motion enabled the investigation of the effects of surface separation on capture and direct comparison to experiment . Many adhesive dynamics simulations aim to discern how more complex factors integrate to influence cellular rolling behaviors ( for example: [32] , [58] , [59] ) . The model presented here did not incorporate cellular factors to try to integrate all of the influences on leukocyte rolling . Rather , the present goal was to answer questions about biomolecular reaction . The effects of surface separation and molecular characteristics governing bond formation on two-dimensional biomolecular kinetics are fundamental questions of biomolecular function . Although the present investigation focuses on the selectins , which are very important to a variety of vascular homing processes , the methodology and results may be applicable to additional classes of two-dimensional bimolecular interactions . Good simulated pausing and skipping results were achieved using physiologic site densities [14] and recently published reaction rate data [60] . Our initial attempt to model rolling behavior using the previously published kf value of 1 . 7 µm2/s [20] did not result in rolling: the sphere formed too many bonds to move . The sphere exhibited good rolling behavior when we employed a kf of 4 . 8×10−4 µm2/s , which was extrapolated from the <AC kfo> recently reported [60] . It is very likely the estimation of the site density was improved in the more recent study . It is also of note that the two different formation rate estimates come from two different measurement methods: the biomembrane force probe and the laser trap . The configuration of the two experiments was different . With the biomembrane force probe study , the two surfaces were held some small distance apart , whereas with the laser trap the two surfaces were pushed together . It is possible an increased confinement of the reactive groups increased the bond formation rate in the measurement with the laser trap . However , it does not seem likely confinement would account for a 3 , 500-fold increase in reaction rate . The simulation method might be employed in the future to investigate the influence of bond formation rates and contour lengths on pausing and skipping behaviors , given the observed sensitivity to the bond formation rate , kf , and vertical transport . A state diagram of their influence may be informative [34] . A direct comparison individually trading each molecular parameter measured for L-selectin and P-selectin should better elucidate the impact of their molecular adaptations in capture and rolling . Although only one model of confinement effects was investigated here , the molecular area confinement model described by ( 2 ) , the simulation can be employed to investigate other functional relations describing the confinement effect . The simplified assumptions of the forces governing z-motion in the simulation may also miss interesting behaviors . An interesting potential result of adding a repulsive layer is that a bond might not simply drive the sphere to the wall as in the present work . Subsequent binding events might ratchet the sphere further into the layer due to the highly damped nature of the vertical diffusion . The recapture time could decrease much more substantially with subsequent binding events . Such an effect might effectively couple an increase in shear rate with an increased force pushing the sphere into the repulsive layer , enhancing recapture with increasing shear . Experimental progress has been made to analyze the near-wall vertical motion of microbeads in low ionic strength solutions using total internal reflection microscopy ( TIRM ) [61] . Future experimental efforts might utilize TIRM methods to analyze more physiologic conditions with higher ionic strength buffers and protein coatings . In silico and in vitro glycocalyx analogues could be constructed [9] . Investigations have found an increase in apparent selectin-mediated cell and microbead capture with increasing shear [29] , [39] , [62] , [63] . Several explanations have been proposed: increased force increases the molecular bond formation rate [39] , [62] , the motion of the two surfaces relative to each other increases reaction through lateral transport [54] , [63] , and an increase in cell flattening with increased shear may enhance tethering [49] . Here , it was discovered recapture is enhanced by vertical transport closer to the wall , independent of cell deformation [51] . The simulation results suggest an additional factor that may contribute to enhanced adhesion under flow conditions . The lack of an enhancement in the bond formation rate due to lateral relative motion transport for the present simulation system is not in conflict with previous computational studies [54] . The physical configuration of the previous system was substantially different . Most significantly , the points where the receptors and ligands were attached to their respective surfaces were free to move in the membrane . A receptor and ligand pair could therefore remain in the contact patch if they happened to find each other as the sphere rolled . The lack of an enhanced effective bond formation rate due to lateral relative motion transport in the present analysis does appear to be in conflict with the conclusion of an experimental study employing immobilized receptors [63] . The conclusion that particle sliding , lateral transport , enhances the binding rates is consistent with the presented experimental scaling data and implies that a lateral transport mechanism governs the formation rate . However , other mechanisms that enhance the bond formation rate and scale similarly with shear and size could also account for the result . For example , inter-particle hydrodynamic interactions can influence vertical transport to the wall [47] , [64] . Notably , the frequency of inter-particle interactions would increase with increasing shear . Vertical cell or bead mixing with the surface might increase with increasing shear by inter-particle interactions . Also , force increases with increasing shear , and other studies have suggested increasing force might increase the bond formation rate [39] , [62] . The analysis developed in Protocol S2 suggests transport in the form of lateral sliding should not enhance formation rates when the receptors and ligands are attached to their respective surfaces by an immobile anchor point . An experimental microbead study suggested molecular bond formation rates might be force dependent [39] . This was a bold assertion given the implications for biomolecular reaction theory . The discrepancy in the conclusions between published microbead studies [39] , [63] suggests the simulation developed here be employed with L-selectin parameters and coupled with experiment . It is likely vertical transport plays an even more significant role in systems with L-selectin than was observed in the present results . L-selectin has fewer repeated subdomains and is shorter than P-selectin , which should enhance the importance of vertical transport . Also , the microbeads used in these experimental studies [39] , [63] were smaller than 10 µm . In Figure 3 , it is apparent the smaller spheres diffuse away from the wall more frequently . Therefore , with these experimental systems [39] , [63] , a more significant effect of the confinement model might be observed . Directly coupling a computational methodology with experimental observations of selectin-mediated particle interactions with TIRM or total internal reflection fluorescence microscopy ( TIRFM ) would facilitate direct observations of vertical fluctuations and a conclusive analysis [65] . If the reaction rate enhancement in both microbead studies [39] , [63] is due to the same mechanism and is truly biomolecular in nature , rather than due to transport , the rate enhancement could be interpreted as a macromolecular version of collision or transition state theory as originally developed for covalent bond formation kinetics . The energy scales involved are interesting . From Figure 3B , the modal velocity for a 6 µm-diameter bead at 50 s−1 is roughly 90 µm/s . At 100 s−1 , near the peak tethering rate in the study [63] , a 6 µm-diameter bead would have a modal kinetic energy equivalent to 0 . 5 kBT . The kinetic energy of the particle could be coupled to the reaction efficiency of the binding pocket through the molecular tether . It has been suggested that apparent increases in the cellular tethering rate with shear [29] may be due to increased flattening of the cell [49] . Cell morphology is much more complicated than the rigid sphere case we have considered here . Sedimentation effects on the microvillus length scale may play a role as cellular protrusions bump into the functionalized surface [17] . Additionally , it has been illustrated here that bond dissociation properties may influence measures designed to test bond association . Another factor could be a small lift effect [66] , [67] . Lift should not be important based on the wall proximity criterion . However , even nano-scale vertical displacements have functional binding consequences due to the molecular nature of the capture events . The molecular loading rates observed in Figure 6 begin near the higher limit of the range employed in the biomembrane force probe study of dissociation pathway switching in P-selectin ( 0 . 02–40 pN/ms ) [22] . It was therefore interesting that there was still an observable effect of the off-rate on the observed long-skip distance in Figure 9A , likely because the bonds were unstressed for a brief period . Interestingly , a study with ( poly ) ethyleneglycol linkers found thermodynamic fluctuations in the molecular tether allowed receptors to extend , bind , and then exert an attractive force between the two surfaces [68] . The assumption that molecular tether points must be brought within the molecular contour length and exert no force upon binding serves as a simplifying first approximation . Measurements of catch-slip bonding do exhibit a striking numerical relationship with previous molecular measurements of length and force . Force at the non-deformed molecular length of 92 nm was 56 pN , as calculated from the freely-jointed chain model parameters [21] . For a dimer , this force distributes as 28 pN per binding head , very close to the reported catch-slip optimum for P-selectin [5] . The peak loading rates in Figure 6 agree with those deduced for neutrophils tethering in a flow cell [29] . Simulations have shown how dissociation of a receptor from the cytoskeleton and microvillus extension can decrease the load on a tethering biomolecular complex , and it was noted clustering plays an important role in cellular bond lifetime [49] . Here , it was observed that in rigid microbead flow assays , the bond loading might be higher but clustering still can explain the discrepancy in the results between measurement methods . Indeed , comparing Figures 7C , E , it might be difficult to make observations of single-molecule bond formation events in a flow cell without a careful experimental design . Although the present study suggests molecular confinement is not important to enhancing the function of molecular pairs mediating transient interactions , confinement has been suggested to play a role for molecular pairs that must mediate longer-lived interactions [7] . Although flow cell techniques have frequently been used to investigate interactions involving selectins , they have been applied to more classes of molecules such as antibodies , cadherins , and T-cell surface molecules [69]–[71] . The wobble autocorrelation measurement should be broadly applicable to more classes of molecular interactions than the P-selectin/PSGL-1 interaction explicitly explored here , where it might better indicate adhesive function . The implementation of a new modeling methodology to investigate the important qualitative and quantitative characteristics of molecular systems mediating two-dimensional interactions was reported here . In addition to exploring the molecular characteristics and parameters important for other closely related adhesion systems , such as L-selectin/PSGL-1 , we anticipate the computational method will be extended to entirely new molecular systems . For example , polyethylene glycol tethers have a pronounced influence on particle interactions with immobilized ligands [72] . Properties such the effective tether extension and the compressibility of the surrounding polyethylene glycol coat might be designed using computational modeling for optimized vascular binding to molecular targets . Although the present focus has been on understanding dynamic interaction , e . g . biomolecular tethers arresting particles , the technique should be entirely applicable to new classes of molecules . The investigation of confinement was inspired by a study of CD2/LFA3 interaction , an important component of extended adhesion and signaling between T-cells and antigen presenting cells [7] . The technique should therefore be broadly applicable to additional classes of inter-cellular interactions . The importance of confinement in mediating long-lived interactions has been suggested to be a result of the closely-controlled intermembrane distance , which is not fixed in dynamic adhesion . For example , the diffusion of individual receptors might be added into a discrete receptor model to watch how molecules assemble into the synapse at the interface during the extended adhesive interaction . Furthermore , the computational methods might be used to optimize the molecular properties , such as length and flexibility . Intercellular bond formation could be linked to intracellular signaling cascades and the spatial localization of signaling scaffolds . This could facilitate the design of functionally-enhanced dendritic cells for immunotherapy or re-engineering dendritic cell subpopulations to elicit a desired T-cell ( usually TH1 ) differentiation pathway [73] . | The binding of a receptor on one cell to a ligand on another is a process of broad biological interest , important to cell adhesion and signaling . Interactions between cell surfaces can be called “two-dimensional” because the reactive groups on interacting molecular pairs are constrained to move 100 nm or less in the direction perpendicular to the surfaces . The molecular reactive groups are attached to their respective cellular surfaces through a molecular tether embedded in the cell membrane . There are many parameters that might affect the observed binding kinetics , such as the distance between the cell surfaces , the length of the molecular tether , and the freedom of the reactive groups to move about on their molecular tether . A well-studied case of two-dimensional interactions is that through which circulating leukocytes capture to the endothelium and exit the blood into the tissues . Leukocyte capture presents an additional complexity: bonds that restrain leukocytes against the shearing force exerted by the blood must be capable of withstanding the force trying to pull the receptor and ligand apart at their noncovalent interface . New models have been proposed to explain the behavior of individual receptors and ligands , raising the question: which molecular behaviors have an effect on function ? | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biophysics/biomacromolecule-ligand",
"interactions",
"cell",
"biology/cell",
"adhesion",
"biophysics/theory",
"and",
"simulation"
] | 2009 | Nano-motion Dynamics are Determined by Surface-Tethered Selectin Mechanokinetics and Bond Formation |
A steady increase in knowledge of the molecular and antigenic structure of the gp120 and gp41 HIV-1 envelope glycoproteins ( Env ) is yielding important new insights for vaccine design , but it has been difficult to translate this information to an immunogen that elicits broadly neutralizing antibodies . To help bridge this gap , we used phylogenetically corrected statistical methods to identify amino acid signature patterns in Envs derived from people who have made potently neutralizing antibodies , with the hypothesis that these Envs may share common features that would be useful for incorporation in a vaccine immunogen . Before attempting this , essentially as a control , we explored the utility of our computational methods for defining signatures of complex neutralization phenotypes by analyzing Env sequences from 251 clonal viruses that were differentially sensitive to neutralization by the well-characterized gp120-specific monoclonal antibody , b12 . We identified ten b12-neutralization signatures , including seven either in the b12-binding surface of gp120 or in the V2 region of gp120 that have been previously shown to impact b12 sensitivity . A simple algorithm based on the b12 signature pattern was predictive of b12 sensitivity/resistance in an additional blinded panel of 57 viruses . Upon obtaining these reassuring outcomes , we went on to apply these same computational methods to define signature patterns in Env from HIV-1 infected individuals who had potent , broadly neutralizing responses . We analyzed a checkerboard-style neutralization dataset with sera from 69 HIV-1-infected individuals tested against a panel of 25 different Envs . Distinct clusters of sera with high and low neutralization potencies were identified . Six signature positions in Env sequences obtained from the 69 samples were found to be strongly associated with either the high or low potency responses . Five sites were in the CD4-induced coreceptor binding site of gp120 , suggesting an important role for this region in the elicitation of broadly neutralizing antibody responses against HIV-1 .
Elicitation of broadly cross-reactive neutralizing antibody ( NAb ) responses is a high priority for HIV-1 vaccines [1]–[4] . Many candidate immunogens elicit strong NAb responses against highly neutralization-sensitive strains of HIV-1; however , these vaccine-elicited antibodies neutralize very few circulating strains [5]–[7] and have not afforded protection in past human efficacy trials [8]–[10] . A recently completed efficacy trial in Thailand ( RV144 ) , in which a modest reduction in the rate of HIV-1 infection was observed [11] , provides hope that with further improvements a more acceptable level of efficacy is obtainable . It is too soon to know whether NAbs contributed to the observed efficacy in RV144 . Based on immunogenicity data from earlier phase I and II clinical trials of this and related vaccines [4] , [12] , improved NAb responses may be one way to achieve greater protection . Such improvements are likely to require novel vaccinedesigns . Most current efforts to design NAb-based HIV-1 vaccine immunogens are guided in part by knowledge of the molecular structure of the viral Envelope ( Env ) glycoproteins that serve as the sole targets for NAbs [13]–[16] . These Env glycoproteins consist of a surface gp120 and transmembrane gp41 that associate non-covalently and assemble into a trimeric complex of gp120-gp41 heterodimers on the virus surface , where the mature Env trimer spike mediates virus entry into host cells [17]–[19] . Entry is mediated by successive binding of gp120 to its cellular CD4 receptor and an obligatory coreceptor , most often the chemokine receptor CCR5 , triggering conformational changes that permit gp41 to induce membrane fusion [18]–[20] . Env trimers and their individual constituents are genetically variable , conformationally flexible and heavily glycosylated , making them difficult targets for NAbs [1] , [2] , [19] , [21] . Because fitness constraints do not permit the virus to evolve to become completely resistant to neutralization [22] , [23] , certain NAb epitopes remain vulnerable that are of particular interest for vaccine development . Some of these epitopes are well studied , whereas others remain unknown or only partially characterized [2] , [4] , [24] . The structural complexity of Env requires sophisticated methods for the analysis of NAb epitopes . X-ray crystallography and cryo-electron tomography , together with data from mutagenesis and biophysical studies , have been used to illuminate several vulnerable regions in great detail . Examples of how this information is used for novel immunogen designs include the optimization and stabilization of epitopes in the receptor and coreceptor binding regions of gp120 [25]–[27] . Other examples include innovative structural variants of gp41 [28]–[30] and optimal mimics of gp120 and gp41 epitopes recognized by broadly neutralizing monoclonal antibodies ( mAbs ) [31]–[35] . Although these new design efforts are in early stages of testing , none so far have yielded substantial improvements . Many new concepts for NAb-inducing vaccines based on HIV-1 Env are being explored . These concepts are complicated by inconsistencies between the antigenic and immunogenic properties of key epitopes . For example , Env antigens that possess high affinity epitopes for broadly neutralizing mAbs fail to elicit these types of antibodies [28]–[30] , [36]–[39] . Also , gp120 antigens similar to those that performed poorly as early vaccine candidates contain epitopes that are capable of absorbing-out a substantial fraction of broadly NAbs in sera from a subset of HIV-1-infected individuals [40]–[43] . Some B cell responses might be down regulated by self-tolerance mechanisms , as has been suggested for epitopes in the membrane proximal external region ( MPER ) of gp41 [44] , [45] . Other B cell responses might be down regulated by immunosuppressive properties of gp120 [46]–[48] . Although it remains unclear why some of the most attractive Env epitopes are poor immunogens , the potent neutralizing activities of a subset of human mAbs [49] , [50] and sera from HIV-1-infected individuals [51] suggest it might be possible to design better vaccine immunogens . A greater understanding of the antigenic and immunogenic properties of Env should facilitate the discovery of an effective HIV-1 vaccine . We ( and others ) are using computational analyses of large neutralization datasets derived from assays with HIV-1-positive sera and molecularly cloned Env-pseudotyped viruses to gain new insights . Statistically significant associations are sought between amino acids in particular positions in the alignment and either i ) the neutralization susceptibility of a given Env , or ii ) the potency and cross-reactivity of neutralizing antibody responses of individuals harboring a given Env . Here we use the term “signatures” to refer to the amino acids in a given position in Env that are associated with a neutralization phenotype . Previously , several amino acid signatures in gp120 and gp41 were identified that strongly associate with the antigenic determinants of NAbs in sera from HIV-1-infected subjects [52] , [53] . Such signatures could either be a consequence of direct contacts for NAbs , or reflect conformational requirements/constraints that regulate Ab access . Because of the distinctive lineages in HIV evolution , found at multiple levels ( clades , subclades , and geographic clusters within a clade ) , it is critical to correct for the phylogenetic associations among sequences when defining signatures rather than merely attempting to predict phenotypes . Not accounting for phylogeny can lead to spurious positive signals that result from lineage effects and a reduced sensitivity , as was seen when associations were sought between host HLA and HIV amino acid substitutions at the population level [54] , [55] . We used three distinct phylogenetically corrected statistical approaches to look for signature patterns . The first was the approach taken by Bhattacharya et al . [54] , but modified to enable looking at combinations of sites and combinations of amino acids within sites . The two novel statistical strategies for defining signatures used here are conditional mutual information and a modified decision forest approach . We first tested our computational signature identification methods in the context of neutralizing antibody signatures by accurately identifying a subset of the known determinants of the epitope for broadly neutralizing mAb b12 . Here our primary goal was not prediction of the phenotype of unknown sequences; rather the main goals were to identify amino acid mutational patterns that correlate with b12 sensitivity independently of founder effects , and then form hypotheses regarding sites/mutations that may directly impact neutralization , to use biological knowledge available in the literature to evaluate these hypotheses , and to suggest further experiments to validate sites/mutations for which knowledge is presently lacking . The b12 signature patterns we identified were well supported by the literature , indicating the computational methods were indeed identifying meaningful sites . We then applied these methods in a reciprocal fashion to determine whether amino acid signatures in the Env proteins from HIV-1 infected individuals with particularly broad NAb responses could be identified relative to individuals who do not elicit broad responses . Our hypothesis was that there may be common features in Envs capable of eliciting potent neutralizing antibodies , and identification of such signatures may ultimately be helpful for immunogen design . Our findings suggest that broadly NAb responses are determined in part by features in the CD4-induced ( CD4i ) co-receptor binding site ( CoRbs ) of gp120 .
Neutralization data and Env sequences relating to the b12 epitope that overlaps the CD4 binding site ( CD4bs ) of gp120 [35] were analyzed as a means to partially validate our computational methods for signature site identification . The mAb b12 was chosen for methodological validation purposes both because many details regarding its epitope are known , and because it is an epitope of great interest for vaccine design . The analyses utilized genetic sequences and b12 sensitivities of 251 clonal Env-pseudotyped viruses representing many HIV subtypes , recombinant lineages and disease stages ( Fig . 1 , Table S1 ) . IC50 values were determined from neutralization curves where the highest dose of b12 tested was either 25 µg/ml or 50 µg/ml , depending on the experiment . Viruses not neutralized at the highest dose tested are referred to here as being resistant; that is not to say , however , that some of the viruses would not have been neutralized by higher b12 concentrations . Among the 251 viruses tested , 88 ( 35% ) were sensitive at varying levels ( Table S2 ) , and the other 163 were resistant at the highest concentration tested . First , other potential correlates of b12 sensitivity were examined , including viral genetic subtype , sensitivity to soluble CD4 ( sCD4 ) , and the disease stage of the donor at the time of virus isolation . Multiple subtypes were included in the study ( Fig . 1 ) . Envs that were B subtype exhibited the highest frequency of b12 neutralization susceptibility ( Fisher's exact test p = 3 . 6×10−4 , comparing B subtype to all others , Fig . 2A ) . In situations like this , in which there is a strong clade structure in the evolutionary tree and an enrichment of the phenotype of interest in a particular clade , it is critical to employ strategies that include phylogenetic correction to avoid spurious positives when seeking amino acid signatures . This is because it is likely that only a small number out of the amino acids that are commonly enriched in the B clade will directly impact an Envelope's susceptibility to b12; however , any amino acid enriched in the B clade , including amino acids that are common in the B clade due to founder effects , will be biased towards appearing associated with b12 sensitivity ( Fig . 1 ) . Envs of the target viruses were obtained and sequenced at different stages of infection . The Fiebig stage [56] for most subjects at the time the Env was sampled was experimentally determined as an indicator of stage of infection ( Table S1 ) . When the Fiebig stage was not experimentally determined , the subjects were generally noted to be in a “chronic” or “early/acute” stage at the time the sample was obtained ( Table S1 ) . When the subjects were broken into categories of “chronic” ( grouping those in Fiebig stages VI or V/VI , with those noted to be in chronic infection ) and “early” ( grouping Fiebig stages I-V , with those noted to be in acute or early infection ) there was no difference between b12 sensitivity or resistance , nor was there any correlation between b12 sensitivity and the series of Fiebig stages ( data not shown ) . Thus the results from our cross-sectional examination of b12 resistance at different stages of infection suggests that the emergence of b12 resistance over time that was previously observed in a longitudinal study in a small number of subjects [57] may not be a common pattern . Finally , consistent with previous findings [58] , Envs that were susceptible to b12 neutralization were more sensitive to neutralization by sCD4 ( p = 0 . 0001 , Wilcoxon rank sum test , Fig . 2B ) . Among just the b12 sensitive viruses , there was a weak correlation between the neutralizing potencies of b12 and sCD4 ( Kendall tau Rank Correlation: p = 0 . 0015 , tau = 0 . 23 , data not shown ) . Our signature analyses strategies identified ten b12 sensitivity amino acid signatures in Env . Associations with a q value ( false discovery rate ) <0 . 2 are presented in Tables 1 and 2; a relatively high q value cut off was used to be inclusive at this hypothesis forming stage . Seven signatures ( 6 in gp120 and 1 in gp41 ) were identified by phylogenetically corrected contingency table analysis [54] ( example shown in Fig . 1 ) . Specific amino acid mutational patterns in each position formed the basis of contingency table analysis; these are noted in Tables 1 and 2 . We used several different likelihood trees as input to test the sensitivity of the signature analysis results to the phylogenetic tree . Two distinct tree topologies from two different runs on parallel computers ( see Materials and Methods section ) gave identical signature results in terms of sites and amino acids , with , as expected , slightly different p and q-values . A tree generated using PHYML [59] , a program that can be run very quickly but had a less optimal likelihood scores than our trees based on a more extensive exploration of the tree topologies , identified 6/7 of the original signature sites , but missed the site 651 ( which we have since shown experimentally to impact b12 sensitivity , see below ) , and captured two other sites ( site 364 and 742 ) with very borderline q-values ( ∼0 . 19 ) . Given the intrinsic variation in the trees , and our inclusive high q value cutoff of 0 . 2 for hypothesis generation , one would expect some run-to-run variation . The overall consistency of the signature results based on the 3 trees , however , suggests the results are relatively robust and independent of the tree; we present in Tables 1 and 2 the sites and statistics based on the tree with the maximum likelihood . A representative example of a single amino acid contingency analysis through the maximum likelihood tree , Aspartic Acid ( D ) at position 185 , is illustrated in Fig . 1 . The simple uncorrected Fisher's exact p value for this signature amino acid example ( p<10−8 ) indicated that a D in position 185 is highly associated with b12 sensitivity . The low p-values for the patterns of change and stability relative to the most recent ancestral state as estimated through the maximum likelihood tree , showed that mutation away from D in resistant viruses ( p = 0 . 0005 ) , and towards D in sensitive viruses ( p = 0 . 0004 ) were also associated with b12 sensitivity , providing assurance that the profound association with 185D and b12 sensitivity was not simply an artifact of shared lineages ( Fig . 1 , Tables 3 and 4 ) . The low q values ( Table 2 , q = 0 . 06 and q = 0 . 04 , respectively ) indicate that these low p-values are not expected by chance alone , despite the very large number of tests performed ( i . e . , every amino acid found in every position in Env , and all combinations of amino acids in every position ) . We also analyzed all potential N-linked glycosylation sites ( based on the presence or absence of the amino acid N-linked glycosylation motif NX[ST] in a give position in the Env alignment ) for associations with b12 activity , again using a phylogenetic correction . None had a q-value<0 . 2 , and the only one that showed borderline significance was found at position 149 ( it is noted in Table 2 ) . Finally , we also explored the b12/gp120 interface more deeply , including all combinations of amino acids in all pairs of sites in this region . Single sites accounted for most of the statistically significant signatures in the b12 binding region ( Table 2 ) . ( A listing of the sites included in the b12 binding region is available in Table S3 ) . Of these 7 sites defined by phylogenetically corrected contingency analyses , 5 were also identified as b12 signatures by an ensemble learning technique using classification trees , while 3 were also identified by conditional mutual information ( CMI ) analysis ( Table 1 ) . The best predictors from the ensemble learning approach included a subset of the most significant amino acids in the contingency table ( Tables 1 and 2 ) , and did not add any new information . An additional 3 signature sites were uniquely identified by CMI analysis: 2 in gp120 and 1 in gp41 ( Tables 1 and 2 ) . The CMI approach was used to increase our sensitivity , and possibly to capture additional sites of interest . The contingency table analyses restrict each comparison at each site to a particular amino acid or combinations of amino acids in the ancestral states immediately preceding the endpoint taxa , thus using only a subset of the available data for statistical analysis . In contrast , CMI utilizes information across all possible ancestral states at the immediate ancestral node of the tip , but does not identify particular amino acids at the site of interest , just the sites that had mutational patterns associated with resistance or susceptibility . An alignment of the three additional sites that were identified by the CMI method is provided in supplement Fig . S1 . Each of these positions was relatively conserved; examining these alignments suggests the consensus amino acids at the three sites , 163T , 182V , and 655K , are well tolerated among viruses with b12 sensitivity , but that mutations 163A , 182E and mutations away from 655K , were enriched among resistant viruses . It is important to remember that while these associations are statistically supported ( Tables 1 and 2 ) , any mutation in isolation may not be able to alter the phenotype of a virus in the context of a given natural strain . For example , although a change away from D at position 185 was most significantly associated with b12 resistance , and was most predictive of the phenotype , some Envs carrying the mutation remained b12 sensitive in 13/48 ( 27% ) natural occurrences of this pattern . Thus , the signatures we identified point to the biological relevance of mutational patterns among a population of circulating viruses , but are not necessarily predictive in isolation in a single strain . Despite this , higher frequencies of amino acid substitutions associated with a b12 resistant phenotype , and loss of substitutions associated with a b12 sensitive phenotype , summed over all 7 signature sites , were strongly associated with resistance . This indicates that effects at the positions identified were cumulative . Notably , the signature sites were identified based on a simple Boolean resistant/sensitive phenotype , yet resistance-associated amino acids accumulated across these sites in viruses with diminishing b12 sensitivity . Specifically , the left hand box in Fig . 3 includes all b12 sensitive pseudoviruses tested , and is ordered by diminishing sensitivity . Combinations of more resistant and fewer sensitive amino acids are clearly evident among the least sensitive viruses nearing the end of the columns . The combined results of the contingency table signature analyses identified five statistically significant signature sites that resided in , or proximal to , the CCR5 CoRbs of gp120 ( Tables 5 and 6 ) . These sites are shown in a crystallographic model of gp120 complexed with CD4 and the CD4i-specific mAb 17b in Fig . 8 . Sites 419 and 421 are located in the V4 region of gp120 , immediately adjacent to the β20 strand of the bridging sheet that connects the inner and outer domains of gp120 [35] , [61] . Both sites make contact with the CD4i-specific mAb 17b [61] ( Fig . 8 ) and have been shown to be critical for CCR5 co-receptor binding [95]–[98] . Site 419 also makes contact with b12 [35] , whereas site 421 is involved in the binding of other CD4i-specific mAbs E51 [98] and 48d [99] as well . Sites 413 and 440 in V4 and C5 , respectively , are spatially close to the bridging sheet and overlap the contact surface for 17b [61] . Site 440 has been shown to be critical for CCR5 binding [96]–[98] . CMI analysis identified an additional site in the V2 loop , position 186 , immediately adjacent to the b12 signature site at position 185 . In addition to the position-based signature analysis , we found that strong NAb responses were associated with serum Env proteins that had fewer PNLGs and shorter lengths in V2 ( Table 6 ) . It has been shown that V1/V2 stem region can impact CCR5 binding since it plays a significant role in formation of the bridging sheet [96] , [97] . Furthermore , site-directed mutational studies have shown that regions outside V3 loop , including site 166 ( a position within V2 loop ) can play a significant role in co-receptor usage/switch [94] , [100] . Considering the flexibility of the loop and ensuing conformational changes that take place involving V1/V2 upon CD4 binding , a position such as 186 can directly or indirectly interact with critical sites involve in the formation of bridging sheet . The fact that no other signatures were identified suggests that the CCR5 CoRbs may play a substantial and relatively consistent role in the NAb response in HIV-1-infected individuals .
Assay technologies that utilize molecularly cloned Env-pseudotyped viruses with a defined sequence provide powerful tools for dissecting molecular determinants of neutralization epitopes on HIV-1 . In addition to enabling mutagenesis studies , data from assays with clonal Env-pseudotyped viruses have been used for computational analysis to identify Env amino acid signatures that associate with the antigenic recognition patterns of autologous [53] and heterologous [52] NAbs in sera from HIV-1-infected individuals; such signatures could be contact sites for NAbs , or they may be determinants of epitope exposure in the quaternary structure of Env spikes . Here we obtained partial validation of a computational strategy to identify amino acid positions that are related to NAb phenotypes . We systematically studied patterns of mutations in Env proteins that correlate with b12 susceptibility , and our signature analysis successfully identified key positions that are known from crystallographic and mutagenesis studies to be critical sites in the b12 epitope . Thus , 7/10 signature sites were identified either directly in the contact surface for b12 , or in V2 , which is known to impact b12 binding and neutralization . Notably , mutations in position 185 in V2 were nearly equal in strength to mutations in position 461 , which were in this study the two best predictors for assessing b12 neutralization susceptibility in natural strains . Three novel positions were implicated in b12 neutralization based on amino acid associations . One was in gp120 , at position 268; this signature raised a plausible hypothesis regarding the impact of electrostatic potential at the isosurface of gp120 on interactions with the positively charged b12 antibody . Two additional b12 signatures were identified in gp41 that were intriguing because they may affect exposure of the b12 epitope in the quaternary structure of Env . Interestingly , both sites in gp41 directly co-varied with sites at the b12-gp120 interface . Site-specific mutagenesis studies have been initiated to explore the impact of positions 268 and 651 on b12 neutralization , and the predicted signature substitutions at these positions were indeed found to capable of significantly impact b12 binding ( manuscript in preparation ) . Two of the ten sites identified are statistically expected to be false positives , so it is likely that two will be not be found to be relevant when experimentally tested , although each of the ten sites and amino acids associations are biologically plausible , and most already validated in the experimental literature , as discussed in the results . Thus the b12 analysis provided a validation of using this approach for identifying signature patterns related to neutralization phenotypes , and provided new information by defining the particular mutations in the natural virus population that correlate with b12 neutralization susceptibility , and by determining the relative strength of such associations ( Table 2 ) . The apparent accuracy of the b12 susceptibility signature analysis was encouraging; however , our findings highlight both limitations and virtues of these methods . Sequence-based signatures methods cannot be expected to identify all b12 contact residues in gp120 [35]; this is because some of these sites are highly conserved , whereas other sites at the contact interface may have natural variation that is well tolerated by b12 . Yet other important sites might reside in hypervariable regions that cannot be aligned with confidence , so were excluded from our analysis – we attempted to examine the impact of these regions based on loop lengths and total number of glycosylation sites , which are alignment independent measures . In addition , since these methods scan full Env and started with no biological priors , they necessarily are based on a large number of tests that makes detecting weak signatures prohibitively data intensive . For example , we did not identify two PNLGs known to affect b12 susceptibility [58] . One of these sites was at the base the V2 loop ( position 197 ) and the other was in the V3 loop ( position 301 ) . The PNLG in position 197 is almost invariant , and so could not have been identified by our method , which relies on detecting associations in the context of natural sequence variability . Position 301 ( PNLG ) reached borderline significance in the complete scan of Env when testing for an association between the preservation or loss of specific PNLGs and the b12 neutralization ( p = 0 . 019 , q = 0 . 30 , OR = 0 . 23 ) . Signature methods focus on sites that are likely to be the most impacted by common mutational patterns found in the circulating population . Such mutational patterns are directly relevant for vaccine design considerations because we must contend with the natural variation of HIV for a vaccine to succeed . Indeed , signature methods provide a useful counterpoint to crystallography , which identifies the contact surface of a protein bound by antibody , but does not provide direct information about the implications of key common natural mutations [35] . Moreover , alanine scanning [64] , which explores the functional impact of mutations introduced in either conserved or variable positions , is another extremely valuable tool , but one that is limited in terms of being able to look at the consequences of natural variation at specific sites or in combinations of sites . An additional limitation is experimental , in that some sites might require concentrations of b12 that are higher than those used here for positive identification . Despite these limitations , our computational analysis appears useful for delineating molecular determinants of complex neutralization epitopes on HIV-1 Env , including the identification distant sites that may impact b12 binding though quaternary and allosteric effects . The neutralizing impact of b12 is very specific , where slight differences in recognition sites between viruses can have major phenotypic consequences [101] . A better understanding of the impact of common natural mutations that are outside of the immediate binding surface of b12 may ultimately allow improved rational design strategies of vaccines that attempt to elicit potent anti-CD4bs antibodies . Having confirmed that our computational analysis has utility for identifying molecular determinants of Env antigenicity in the context of the b12 epitope , we then sought to determine whether a similar computational analysis , but this time based on Env sequences derived from serum samples from HIV-1-infected individuals , could identify amino acid signatures that associate with the magnitude and breadth of the neutralizing activity of the serum samples . Any signatures identified by such analysis might be determinants of the immunogenic as well as antigenic properties of Env , although it was beyond the scope of this study to discriminate between these two immunologic properties . For the analyses of Env sequences in serum samples that were evaluated for neutralizing activity , a single Env sequence from each individual was obtained . We were interested in leveraging our resources to increase the number of individuals studied rather than increasing the depth of characterization of each infected individual . In part we were testing the feasibility of the approach for scanning a large population of HIV infected individuals with the intent of finding common features of the virus harbored in them that may have given rise to a potent NAb response . Viral evolution and quasispecies complexity in chronically infected subjects clearly were potential confounding factors; the single sequence used was randomly selected from a complex viral population within each individual and may not reflect the form of the Env that gave rise to the NAbs of interest in the serum samples . Indeed , assuming that the NAb response during chronic infection is driven by multiple viral variants , these confounding factors limit our ability to identify genetic signatures . Despite this , statistically significant signatures were revealed based on an analysis of sequences from a single Env clone from a single time point from each of 69 individuals , indicating a detectable consistency of signal across the population . Notably , despite scanning the full Env , these signatures were focused on a single biologically interesting region , the CoRbs . An unresolved issue that is an inherent consequence of this signature-defining strategy is the uncertainty regarding whether the signature amino acids reflect common features that were useful for stimulating potent NAb responses , or if instead they reflect common patterns of escape from the NAb responses in the potent sera . Experimental comparisons to resolve this are underway; strains that retain the signature positions that are associated with potent sera , like CH0219 . e4 and CH080510 . e . p2 ( Fig . 7 ) , are particularly interesting candidates for immunogenicity testing . The fact that five of the six signature sites identified , with one false positive expected , were in the CoRbs of gp120 suggests an important role for this region in generating high titers of broadly NAb responses . This region is comprised of elements of the bridging sheet and adjacent surfaces from the outer domain of gp120 , including the V3 loop , that undergo conformational changes and become exposed upon CD4 binding as an intermediate step in the membrane fusion process [61] , [96] , [97] , [102]–[104] . It is possible that in some cases CD4i-specific mAbs contribute directly to potent cross-neutralizing ability [87] , [95] . The CoRbs is one of the most highly conserved and protected domains on gp120 [86] . Rare variants of HIV-1 exist that exhibit spontaneous exposure of CD4i epitopes; these strains tend to infect cells independently of CD4 and to be highly sensitivity to neutralization by CoR-specific antibodies [105] , [106] . Owing to the presence of such antibodies in HIV-1-infected individuals [86] , [87] , [95] , a mechanism of CD4-induced exposure of the CoRbs serves as an effective strategy to evade humoral immunity — a strategy that is aided by steric constraints that prevent anti-CoR antibodies from gaining accessing to their epitopes at the virus-cell interface [107] . In a systematic thermodynamic analysis by Kwong et al . , in which 20 antibodies were categorized according to where they bind on the gp120 surface , it was found that 6 of 7 antibodies that bind gp120 at its receptor and coreceptor binding sites exhibited unusually high binding entropy ( including 17b that binds to CoRbs ) [21] . Therefore , the signature sites identified here in the CoRbs might play an indirect role in neutralization by antibodies that induce large conformational changes in gp120 . The question naturally arises as to why a region of gp120 that is so heavily guarded and difficult to target by NAbs registered in our analysis as a key determinant of potent NAb responses in HIV-1-infected individuals . One possibility is that the CoRbs of gp120 has vulnerabilities that are only beginning to be recognized . For example , using a novel combination of epitope mapping techniques , Li et al . [95] reported evidence that CoRbs-specific antibodies contributed to the broadly cross-reactive neutralizing activity of serum from two HIV-1 infected individuals . In addition , CoRbs residues were implicated by alanine scanning mutagenesis as being involved to a minor extent in the epitopes for two newly described broadly neutralizing mAbs [50] . Also , vaccine-elicited CoRbs-specific antibodies correlated with viremia control in a simian-human immunodeficiency virus ( SHIV ) challenge model in nonhuman primates [108] . It also seems possible that amino acid residues in key positions in the CoRbs of gp120 modulate the conformation of adjacent regions , such as the CD4bs , much the same as conformational changes induced by gp120-CD4 binding modulate the CoRbs . Limited sequence variability in the CD4bs [109] , [110] makes this an attractive target for NAb-based vaccines . Indeed , studies have shown that the CD4bs is targeted by broadly NAbs in sera from some HIV-1-infected individuals [51] . It remains to be determined whether the genetic signatures of potent NAb responses identified here contribute to the immunogenicity as well as antigenicity of Env . By design we were attempting to resolve signatures that impacted Env immunogenicity in natural infection . Clearly , strong antigenicity alone is generally not sufficient for the elicitation of NAbs [28]–[30] , [36]–[39] . Other requirements may need to be met before B cells can be stimulated to produce NAbs against certain epitopes of interest . Although very little is known about what these requirements might be , proper Env configuration for B cell recognition and antibody affinity maturation should be considered . It will be interesting to test novel Env immunogens that naturally contain the genetic signatures identified in our study , or that introduce these signatures experimentally . At the very least , our findings suggest that greater attention should be paid to the CoRbs of gp120 when designing novel vaccine immunogens .
All viruses were used as molecularly cloned Env-pseudotyped viruses that expressed the entire gp160 of the designated strain . The multisubtype panel of viruses used for analysis of b12 neutralization is described in Tables S1 and S2 . The 25 viruses used to assess the neutralizing activity of HIV-1-positive serum samples were isolated from sexually acquired infections and were sampled early in infection to closely resembled transmitted/founder viruses . Among these , isolates 6535 . 3 , QH0692 . 42 , SC422661 . 8 , PVO . 4 , AC10 . 0 . 29 and RHPA4259 . 7 belong to a recommended panel of subtype B reference strains [111] . Isolates Du156 . 12 , Du172 . 17 , Du422 . 1 , ZM197M . PB7 and ZM214M . PL15 belong to a recommended panel of subtype C reference strains [112] . Isolates Q23 . 17 , Q842 . d12 , Q168 . a2 , Q259 . d2 . 17 , Q461 . e2 and Q769 . d22 are subtype A reference strains [113] . Isolates BB1006-11 . C3 . 1601 , BB1054-07 . TC4 . 1499 , 700010040 . C9 . 4520 and WEAU-d15 . 410 . 787 are subtype B clones that were confirmed by single genome amplification ( SGA ) and sequencing analysis to be true transmitted/early founder Envs [56] , as were C subtype isolates Ce1086_B2 , Ce0393_C3 , Ce1176_A3 and Ce2010_F5 [114] . These latter 25 viruses utilized CCR5 as their major coreceptor and were considered to possess a tier 2 neutralization phenotype [115] . Serum samples were obtained from HIV-1-infected subjects who were enrolled in clinical protocols of the Center for HIV/AIDS Vaccine Immunology ( CHAVI ) . All subjects were chronically infected at the time of enrollment . The precise length of time of infection was not known . The mAb b12 was provided by Quality Biologicals , Inc . ( Gaithersburg , MD ) as a complete IgG molecule . The SGA methods used here were described previously [116] and result in sequences that are not corrupted by recombination during amplification . Viral RNA was prepared from 400 µl of patient plasma and eluted into 60 µl of elution buffer using EZ1Virus Mini Kit V2 . 0 ( Qiagen , Valencia , CA ) . Viral cDNA was prepared with 20 µl of vRNA and 80 pmol of primer 1 . R3 . B3R ( 5′-ACTACTTGAAGCACTCAAGGCAAGCTTTATTG-3′ ) in a 50 µl volume using Superscript III ( Invitrogen; Carlsbad , CA ) . SGA of the cDNA was performed using nested PCR to obtain the rev/env cassette and to avoid artificial recombination and resampling of the viral genomes [117] . The cDNA was diluted 1∶3 , 1∶9 and 1∶27 ( 8 reactions per dilution ) to determine a dilution with a positive rate of 20% or less . Each diluted cDNA ( 1 µl ) was used for the first round amplification with primers 07For7 ( 5′CAAATTAYAAAAATTCAAAATTTTCGGGTTTATTACAG-3′ ) and 2 . R3 . B6R ( 5′-TGAAGCACTCAAGGCAAGCTTTATTGAGGC-3′ ) . First round PCR was carried out with 1 unit of Platinum Taq Polymerase High Fidelity ( Invitrogen; Carlsbad , CA ) and 10 pmol of each primer in a 20 µl volume . First round PCR products ( 2 µl ) were used for a second round of PCR with primers VIF1 ( 5′-GGGTTTATTACAGGGACAGCAGAG-3′ ) and Low2c ( 5′-TGAGGCTTAAGCAGTGGGTTCC-3′ ) . The second round PCR used 2 . 5 units of Platinum Taq Polymerase High Fidelity and 20 pmol of each primer in a 50 µl volume . PCR thermocycling conditions were as follows for both rounds of PCR: one cycle at 94°C for 2 minutes; 35 cycles of denaturing step at 94°C for 15 seconds , an annealing step at 60°C for 30 seconds , an extension step at 68°C for 4 minutes , and one cycle at 68°C for 10 minutes . PCR products were visualized on a 1% agarose gel and purified with the QiaQuick PCR Purification kit ( Qiagen; Valencia , CA ) . Sequence analysis of env PCR products was performed on both DNA strands by cycle-sequencing and dye terminator methods using an ABI 3730×l genetic analyzer ( Applied Biosystems; Foster City , CA ) . Individual overlapping sequence fragments for each env SGA were assembled and edited using the Sequencher program 4 . 7 ( Gene Codes , Ann Arbor , MI ) . Subtyping analysis was initially performed using SIMPLOT [118] . All sequences were further validated with RIP and HIV Blast ( www . hiv . lanl . gov ) . Subtyping and recombination discrepancies between the methods were carefully considered and resolved . The single SGA Env sequence obtained from each of the HIV-1 positive individuals with potent or weak neutralizing antibody responses was sampled at random . GenBank accession numbers are provided in the supplementary tables . Neutralization was measured as reductions in luciferase ( Luc ) reporter gene expression after a single round of infection with Env-pseudotyped viruses as described [111] . Briefly , 200 TCID50 of virus was incubated with serial 3-fold dilutions of test sample in duplicate in a total volume of 150 µl for 1 hr at 37°C in 96-well flat-bottom culture plates . Freshly trypsinized TZM-bl cells ( 10 , 000 cells in 100 µl of growth medium containing 37 . 5 µg/ml DEAE dextran ) were added to each well . One set of control wells received cells plus virus ( virus control ) and another set received cells only ( background control ) . After a 48-hour incubation , 100 µl of cells was transferred to a 96-well black solid plates ( Costar ) for measurements of luminescence using the Britelite Luminescence Reporter Gene Assay System ( PerkinElmer Life Sciences ) . Neutralization titers are either the 50% inhibitory dilution ( ID50 , serum samples ) or 50% inhibitory concentration ( IC50 , mAb b12 ) at which relative luminescence units ( RLU ) were reduced by 50% compared to virus control wells after subtraction of background RLUs . Assay stocks of molecularly cloned Env-pseudotyped viruses were prepared by cotransfecting 293T/17 cells with an Env-expressing plasmid and an env-minus backbone plasmid ( pSG3Δenv ) as described [111] . To conduct Env sequence signature analyses with the goal of identifying mutational patterns that correlate with neutralization phenotypes , we first needed to define neutralization phenotypes . For mAb b12 , we initially defined the Envs based on whether or not a 50% reduction in RLU could be achieved at the highest concentration of b12 used; if not , the Env was considered b12 resistant . Some Envs were tested up with to 50 ug/ml of b12 , however others were only tested up to 25 ug/ml ( Table S2 ) ; 9 of the 251 samples had a detectable neutralization response between 25 and 50 ug/ml . Thus did the signature analysis two ways , either treating any result over 25 ug/ml as negative , or treating any positive result as positive; the results were essentially the same either way , and the results presented are based on treating any detected response as positive . This provided a Boolean neutralization sensitive/resistant phenotype to use as a basis for comparing the 251 Envs tested with b12 . Later , we compared the levels of neutralization-sensitivity with the patterns in the b12 signature sites by using IC50 values . Defining a serological phenotype based on a profile of potency of neutralization against a panel of viruses was more complex . We first needed to group HIV-1-positive serum samples that exhibited similar neutralization profiles against a panel of 25 viruses . To achieve this , we used a k-means clustering strategy with two added statistics to assess the robustness of the clusters , factoring in both the uncertainty that results from limited sampling and inter-assay variability ( the impact of experimental noise was explored using a smooth bootstrap ) . Sampling limitations were explored by re-sampling either by rows or columns 1000 times , using a random-with-replacement bootstrap strategy . The impact of inter-assay variation was explored by a smooth bootstrap , re-sampling from a Gaussian model of noise centered on zero and based on a limited number of repeat data values . Noise was adding back to the original scores based on the model . We then re-estimated the k-means clusters 1000 times with noise added back [119] . Using these two strategies we found that no more than k = 3 distinctive clusters of sera were statistically justified , in that 2 or more sera were assigned to each of the three clusters with 90% confidence . Defining more than k = 3 clusters was not justified using this criteria . Sera that were not assigned to a cluster 90% of the time were considered indeterminate; clustering patterns were generally more sensitive to sampling than inter-assay variability . To describe the NAb reactivity pattern of the 3 sera clusters in a Boolean framework ( we are limited to two categories , high versus low ) for signature pattern analyses , we compared Envs that were members of each of the robust serological clusters to all other Envs in the study . For example , we compared the Env sequences associated with the strongest sera ( cluster III ) to the remaining Envs by combining those that were in clusters I and II and those that were poorly resolved . In a second analysis , we set k = 2 and compared just the statistically robust high and low clusters , excluding the intermediate values from the comparison . Alignments used for signature analysis were generated with GeneCutter ( www . hiv . lanl . gov ) , which builds on a HMMER base alignment strategy [120] to provide codon-aligned DNA for phylogenetic and signature analysis . Because hypervariable regions are very difficult to align and compare objectively , we excluded all positions in the alignment that contained more than 10% gaps from the analysis [54] . In practice , this means the difficult to align hypervariable regions and rare insertions were all excluded . Thus , the correlations we find ( listed in Tables 1 and 5 ) are focused in regions of Envelope that are readily aligned , and not a consequence alignment artifacts or impacted by the alignment strategy , with the possible exception of site 186 , which borders a hypervariable region in V2 . Phylogenetically corrected methods were used to identify all signature sites; the contingency table method illustrated in Fig . 1 and Fig . 6 was described in detail in [54] . The reason phylogenetic corrections are critical is that observed patterns in data can result either from correlations imposed by the initial historical emergence of a lineage of viruses ( founder effects ) , or in the case of HIV-1 , a consequence of recent biological interactions . Not accounting for founder effects can lead to erroneous statistical conclusions [52] . While there are many recombinant sequences included in the tree ( as with essentially all HIV population trees ) , limiting the accuracy of the reconstruction of the evolutionary history , the phylogenetic corrections utilized for signature analysis are , however , dependent only on the local region of the tree and the ancestral states near the tips of the branches , reducing the impact of inter-subtype recombination on the analyses . We implement are ancestral reconstruction through maximum likelihood phylogenies , using code originally based on Gary Olsen's fastDNAML [121] , with a GTR model with a likelihood estimate of rate variation per site , and adapted to given ancestral states at all nodes [54] . While a maximum likelihood tree framework enables us to model the ancestral state of the virus immediately preceding the tip , neutralization sensitivity data exists only at the terminal tips of the tree , and we do not attempt to infer phenotypic information at internal nodes . To look for amino acid correlations with phenotype , we tested each position in Env that contained fewer than 10% gaps , for associations of the phenotype of interest based on each amino acid considered alone , any combination of two amino acids at each site , and all combinations of two amino acids at two adjacent sites . We also tested for the preservation or loss of glycosylation sequon motifs at particular positions in the alignment . Finally , we included all possible sets of amino acids in up to 3 positions within four targeted and defined regions of interest ( the structurally defined b12 binding site , the CD4bs , the CoRbs , and the MPER regions , summarized in supplement Table S3 ) . More extensive tests of combinations of sites throughout Env would not have been useful because of power issues given multiple testing . Precisely the same algorithm and series of tests were applied to both the b12 sensitivity data and the data relating to potent neutralization . A large sample size is essential to power explorations of associations between phenotype and mutational patterns . Thus phylogenetic reconstruction can be challenging because the number of possible relationships grows factorially with the number of sequences sampled . To improve our maximum likelihood tree reconstructions , we adapted our phylogenic code [52] to a newly high performance computing platforms ( http://www . lanl . gov/roadrunner/ ) . Trees were run using 25 global rearrangements for up to three days on 512 Cell-accelerated processors , until the likelihood scores no longer improved . The final signature results are relatively robust to the tree however , and a rapidly obtained PHYML tree yielded sound signature results , as discussed in the Results . Access to parallel computing resources at Los Alamos National Laboratory also facilitated protein modeling of loop structures and other computationally intensive and repetitive tasks , such as combinatorial signature analysis and q value calculations . Felsenstein first developed the method of phylogenetically independent contrasts many years ago [122] to address similar problems , i . e . , obtaining phylogenetic corrections when looking for correlations of mutational patterns with quantitative data . We were indeed interested in exploring associations between several quantitative measures and neutralization phenotypes , in particular both loop lengths and the total number of PNLGs in hypervariable regions . However , because hypervariable loop lengths and the number of PNLGs vary rapidly within infected individuals , and so are changing on a time scale much faster then the time scale reflected in the population-based trees , a phylogenetic correction at the population level was deemed not essential in this framework . Thus , for testing the impact of loop lengths and numbers of glycosylation sites , simple Spearman correlation tests were performed . Conditional mutual information ( CMI ) was used as a second computational method to identify positions that exhibit an association between mutation and phenotype ( neutralization sensitivity ) that is independent of phylogenetic lineage . CMI [123] generalizes the conventional mutual information measure [123] that quantifies the association between two objects , e . g . , mutation and phenotype . CMI also quantifies the association between two objects but it conditions the association on a third object , in this case the ancestral state . CMI sums over the associations conditioned on different ancestral state amino acids , and so is potentially more sensitive for detecting associations than the contingency table analysis that involves one ancestor state at a time . On the other hand , if the biological signal exists only for some ancestral states and not others , the extra noise added may reduce the power of the test . The statistical significance of a CMI value at any given position was assessed by fixing the ancestral state to each candidate ancestor state in turn , and permuting the relation between mutation and phenotype 1000 times in order to break any potential association . The distribution of CMI values for such permuted data was used to determine p-values , and q-values were obtained from these using the method of Storey and Tibshirani [124] . As with the Fischer's exact test signatures , a cutoff of q<0 . 2 was used to identify statistically interesting sites , such that a 20% false discovery rate was expected among the identified signatures . To model sequence changes across sites , an ensemble learning technique using classification trees was employed [125] . As with the CMI and contingency table approaches , a sensitive/resistant neutralization category was compared to phylogenetic signals . This neutralization quantity indicates when a virus is neutralized by a fixed amount of b12 antibody . A change observed between an observed amino acid and the corresponding position in the inferred parent sequence provides one phylogenetic signal . Changes toward or away from each observed or inferred amino acid across all of the envelope protein sequences served as the set of phylogenetic signals . Signals are conditioned on the ancestor amino acid; thus , any given position can be an instance of the signal , not an instance of the signal , or not applicable for the signal . To form a decision tree , a signal was first identified that best separated sequences into resistant and sensitive neutralization sets . Each set was then partitioned into two more sets using further signals that best track the neutralization phenotype . This refinement procedure was repeated until no additional signals improved the classification . It was necessary that the classification tree handle the absence of signals as well as their Boolean state in order to avoid phylogenetic artifacts . Prediction was performed by taking a tree and following a main signal , secondary signal , tertiary signal , and so on , according to signal values derived from new data . Even in the absence of mutational signals , a decision tree would still provide a prediction on the basis of whether resistant or sensitive viruses were more common in a training set . It is conceivable that coordinated mutations or reversions could occur in a universal way across viruses ( case 1 ) . Alternatively , the interplay of viruses and hosts could result in different patterns of coordinated sequence change ( case 2 ) . To address the possibility that there can be context dependence on unmeasured quantities ( i . e . , virus behavior groups formed by some unknown process ) , we randomly sampled a subset of the full training data ( 75% ) when building decision trees , performing 140 interactions of decision tree building with different training set samples . 75% of the data was chosen as a trade-off between statistical power ( ability to see any group behavior ) and diversity ( ability to see several groups ) . We chose 140 iterations for computational feasibility . Evaluation of the performance of the decision tree models needed to be separate from the construction of the training data . Thus , before iterating the training set sampling and tree building , we reserved 5 sensitive and 5 resistant viruses for testing purposes . Good models from the 140 decision tree builds were defined as those models that perform better than 60% ( instead of the expectation of 50% for random guesses ) on this reserve dataset . Any one of the 140 training samples and resulting decision trees could represent either case 1 or case 2 , as described above . Therefore , the full process of reserving a random test set and generating 140 models to ‘hit’ each test set was iterated 32 times . For each test set , we obtained on average 10 of the 140 models predictive to at least 60% accuracy . A majority vote of these model predictions was noted for each test set . A “majority vote” was conducted across the 32 test sets to provide the final neutralization prediction . Finally , we identified mutational patterns that recurred most often at the top-level splits in the subset of good models across all runs . These patterns provided another strategy for defining amino acid signatures that correlate with neutralization phenotype ( Table 1 ) , and these were a subset of the sites defined by the most common highest level splits; this method only defined a subset of the statistically promising sites defined by the basic Fischer's exact methods . Unlike other decision forest or bootstrap aggregation approaches ( a . k . a . bagging ) [126] , we cross-validated within the training set and pruned back the trees before using them . This may limit overall accuracy , but it has the advantage that any decision tree model could be interpreted without overtly over-fitting a particular training data set . For structural mapping in gp120 , three different structures were used . We used a structure with loops modeled when residue positions in loops needed to be shown . In this structure , the core of gp120 corresponded to the X-ray structure of CD4-bound YU2 gp120 [127] , with variable loops V1 , V2 and V3 modeled for visualization purpose as described previously [128] . For signature positions in the b12 binding surface of gp120 , we used the X-ray structure corresponding to the PDB code 2NY7 [35] . Finally , for spatial mapping of the signature positions in the CD4i region , we used the X-ray structure with a PDB code , 1RZK , [127] that was solved with a CD4-17b complex . In one instance a three-dimensional structure of gp41 was used to suggest the possibility of allosteric effects within the gp120-gp41 complex . This latter gp41 structure was homology-modeled based on the NMR structure of SIV-1 gp41 structure [129] . Signature positions were mapped onto this structure based on the alignment of sequences with respect to HXB2 . The positional numbering refers to HXB2 . Three-dimensional images were generated using VMD [130] . A holdout set of 56 pseudotyped Envs , for which the b12 sensitivity was known but withheld from the analysis team , was kept aside as a fully blinded test set to determine if we could predict the b12 phenotype of Env-pseudotyped viruses based on either just signature amino acid positions or the ensemble learning strategy across full Env . The training and test set of Envs are included in the phylogenetic tree shown in Fig . 1; viruses known to be b12-sensitive are magenta , those known to be b12-resistant are dark grey , and those used as a blinded test set are light gray . Several strategies to predict phenotype were employed , including the simple requirement of at least 4 sensitive and no more than 1 resistant amino acid in the 7 signature sites , a logistic regression based on the 7 signature sites , and the ensemble learning strategy based on the full Env alignment . A prediction of b12 sensitivity or resistance was made based on all three strategies ( Tables 3 and 4 , Table S2 ) for each of the 251 original training sequences and 56 test sequences . The research presented here was approved by the Duke Institutional Review Board , and the human data was analyzed anonymously using de-identified preexisting samples . | Neutralizing antibodies block infection of cells and thus are considered important to elicit with vaccines . A central problem in HIV-1 vaccine design is that HIV-1 is extremely variable and employs a number of strategies to avoid being recognized by antibodies . Despite this , a subset of infected individuals mounts potent , cross-reactive neutralizing antibody responses . We developed computational strategies for identifying correlations between mutational patterns in the HIV-1 envelope glycoproteins ( gp120 and gp41 ) and neutralization phenotypes . We first applied these methods to define mutations that correlated with susceptibility to the potent neutralizing antibody b12 , as a means to explore the appropriateness of applying our computational strategies to neutralizing antibody phenotypes within the well-understood context of b12-gp120 interactions . Signature sites of known importance were found . We then defined signatures in a panel of envelope glycoproteins sampled from HIV-1-infected individuals who made either potent or weak neutralizing antibody responses , with the hypothesis that common features of the envelope glycoproteins that elicit good antibodies in natural infection might be useful to incorporate as vaccine immunogens . Signature mutations associated with potent neutralizing antibody responses were concentrated in the coreceptor binding site of gp120 – a key region for HIV-1 entry into cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biophysics/structural",
"genomics",
"microbiology/immunity",
"to",
"infections",
"virology/vaccines",
"computational",
"biology/comparative",
"sequence",
"analysis",
"immunology/immune",
"response",
"computational",
"biology/evolutionary",
"modeling",
"infectious",
"diseases/hiv",
... | 2010 | Genetic Signatures in the Envelope Glycoproteins of HIV-1 that Associate with Broadly Neutralizing Antibodies |
Selective routing of information between cortical areas is required in order to combine different sources of information according to cognitive demand . Recent experiments have suggested that alpha band activity originating from the pulvinar coordinates this inter-areal cortical communication . Using a computer model we investigated whether top-down induced shifts in the relative alpha phase between two cortical areas could modulate cortical communication , quantified in terms of changes in gamma band coherence between them . The network model was comprised of two uni-directionally connected neuronal populations of spiking neurons , each representing a cortical area . We find that the phase difference of the alpha oscillations modulating the two neuronal populations strongly affected the interregional gamma-band neuronal coherence . We confirmed that a higher gamma band coherence also resulted in more efficient transmission of spiking information between cortical areas , thereby confirming the value of gamma coherence as a proxy for cortical information transmission . In a model where both neuronal populations were connected bi-directionally , the relative alpha phase determined the directionality of communication between the populations . Our results show the feasibility of a physiological realistic mechanism for routing information in the brain based on coupled oscillations . Our model results in a set of testable predictions regarding phase shifts in alpha oscillations under different task demands requiring experimental quantification of neuronal oscillations in different regions in e . g . attention paradigms .
The selective routing of information between neocortical areas is important for efficient task-specific communication . Since it is impossible for the brain to simultaneously process all the incoming sensory information , it is crucial to enhance the processing of the most relevant information for the task at hand while at the same time ignoring irrelevant or distracting information . This process of selective attention modulates neural activity both in early visual as well as their down-stream areas [1] . Since tasks and goals are represented in higher-order cortical areas , there is a need for a mechanism by which these higher-order areas can influence information processing in lower-order areas . We here explore a mechanism whereby the communication between two lower-order sensory areas is coordinated by a third area inducing phase shifts in slow oscillations . This coordination serves to increase or decrease the efficiency of information transfer depending on task demands . We use the coherence between fast oscillations in these two areas as a proxy for efficiency of information transfer . Synchronization between different cortical rhythms has been suggested as a means for selective communication between cortical areas [2–4] . Experiments in monkeys have shown specifically that synchronization of the gamma rhythm between cortical areas in the occipital cortex is linked to the selective processing of information during visual attention tasks [5] . However , the mechanism by which this selective synchrony modulation of the gamma rhythm is coordinated in a top-down manner remains unknown . One proposal is that the synchronization is achieved by feedforward entrainment , in which a sending region drives the receiving region [6] . There is in addition an abundance of evidence linking the alpha rhythm to the attentional processing of information . Alpha power has specifically been shown to correlate with the level of attentional processing of information [7–10] . When a hemisphere responds to an attended stimulus , alpha power decreases in this hemisphere , lifting inhibition and increasing information processing abilities . At the same time increases in alpha power inhibit processing of information in the other hemisphere . [11–13] . The cortical alpha rhythm is well studied , but its origins remain unclear . There is strong evidence for alpha generators in the infragranular layers of the cortex [14–17] . However , experiments also show that the neocortical alpha activity is coherent with the thalamus [18 , 19] , suggesting that thalamic activity might entrain these infragranular sources . The study by Saalmann and coworkers has linked the alpha rhythm originating from the pulvinar to attentional processing of information in the cortex [20] . The pulvinar has widespread connections to virtually every part of the visual cortex [21 , 22] . When two cortical areas are connected directly by cortico-cortical connections they also receive projections from overlapping populations in the pulvinar [23] . Hence , the pulvinar is an ideal candidate for coordinating the communication between cortical areas . The study by Saalmann and coworkers has shown that selective allocation of attention was associated with an increase in Granger causality ( GC ) from the pulvinar to the parts of the visual cortex that respond to the attended stimuli , lending further support to the idea that the pulvinar coordinates information transmission between cortical areas . Furthermore , this study showed an attention-dependent increase in gamma band as well as alpha band coherence between the relevant parts of cortical areas V4 and TEO , which was correlated with a significant increase of alpha band coherence between the pulvinar and these two cortical areas [20] . In another study , electrical stimulation of the pulvinar caused an increase in firing rate of a cortical neuron when its receptive field ( RF ) overlapped with the receptive field of the stimulated pulvinar neurons , while it lowered the firing rate when the receptive field of the cortical neuron and the pulvinar neurons did not overlap [24] . This indicates that pulvinar projections can enhance or suppress activity in the cortex , possibly influencing cortico-cortical communication . The experimental observation of cross-frequency coupling between alpha phase and gamma band power supports the idea that the alpha rhythm coordinates neuronal processes at higher frequencies [16 , 25] . How this rhythmic modulation of gamma power influences communication remains unknown . As mentioned in the preceding text it has been suggested that increases in alpha power decrease gamma activity in a cortical area , thereby reducing its ability to transmit this information to downstream areas [8] . Another possibility is that the alpha phase modulation of gamma power leads to windows of high gamma power , which can be aligned across different areas to improve transmission . In contrast to the idea that gamma band synchronization is caused by feed-forward entrainment , this would require that top-down modulation of the alpha phase is able to influence gamma band synchronization by adjusting alpha phases in different areas . Experiments conducted to study the role of the alpha phase in cortical communication have focused mainly on determining whether the perception of stimuli depends on the alpha phase at which they were presented [26 , 27] . Recent experiments show that the brain can either actively adjust its alpha phase during an attentional distractor task [28 , 29] or modulate the power [30] . These studies have focused on alpha phase adjustments in tasks with temporal expectations about the onset of stimuli . Although this is often behaviorally relevant , the brain also needs a mechanism to enhance processing of uncued stimuli . A possible mechanism would be to align the alpha phase between different cortical areas rather than adjusting the alpha phase with respect to stimulus onset , thereby making the enhancing effect independent of stimulus onset . Here we investigate by means of a model network comprised of spiking neurons whether the pulvinar could coordinate communication in the gamma band between two cortical areas by aligning the relative alpha phase between these cortical areas . We find that the phase difference between the alpha rhythms of the two populations influences coherence in the gamma band , as well as the amount of stimulus information sent from one area to another . In bi-directional networks we were also able to control the direction of communication by adjusting the alpha phase difference . These results account for a broad set of experimental studies of the role of alpha and gamma oscillations in attentional processing of information .
Consider a network of two connected neocortical areas coordinated by the pulvinar ( Fig 1 ) . To study communication between the neocortical areas we modeled each population as a local network of interconnected regular spiking excitatory ( E ) and inhibitory ( I ) neurons . The inhibitory populations consisted of a combination of fast spiking ( FS ) and low-threshold spiking ( LTS ) interneurons , all modeled using the Izhikevich model [31] with the appropriate parameter setting ( see METHODS ) . Model parameters were adjusted such that the network produced biologically realistic firing rates , specifically 5–10 Hz for E neurons and 25–35 Hz for the two types of I neurons ( see METHODS ) . Gamma oscillations emerged spontaneously through a pyramidal-interneuron gamma ( PING ) mechanism ( [32] ) , in which excitatory input from the pyramidal neurons activated the interneurons . In return they inhibited the network for a period determined by the time scale of fast GABAergic feedback . This resulted in synchronous population activity that oscillated with a frequency in the gamma band ( Fig 2A ) . The oscillations were not regular: the peaks in the spike time histogram ( STH ) as well as the period of the gamma oscillations fluctuated ( Fig 2B ) . The majority of the analyses reported in this paper are performed on the STH signal of the E population . This is because the currents entering and leaving neurons from the E population during synaptic inputs and spiking are thought to contribute most to the extracellular local field potential measured in electrophysiological studies due to their large dipole fields [33] . A clear peak in the power spectrum is observed in the gamma band ( Fig 2C ) . This gamma band is fairly wide reflecting the frequency fluctuations across time . The mean gamma power and frequency strongly depended on the mean input strength to the E and the I neurons ( Fig 2D and 2E ) . The mean gamma power was determined as the average power over a symmetric 20 Hz-wide frequency band centered around the peak frequency , which was separately determined for each individual parameter setting . Increasing input to the E population resulted in an increase in gamma power , while stronger input to the I population reduced gamma power ( Fig 2D ) . The peak frequency of the gamma band increased with stronger input to the E cells as well as the I cells ( Fig 2E ) . These simulations show that we can modulate the power and frequency of gamma oscillations independently by setting the level of depolarization of the inhibitory and excitatory neurons to their appropriate value . To study the coupling between the alpha rhythm and the emergent gamma oscillations we applied a modulatory input of alpha frequency ( 10 Hz ) with an amplitude of 23 pA to the I population as described in the METHODS section . There was no alpha frequency input current to the excitatory neurons , hence the effect of the modulation on the E population was assumed to be indirect and mediated by the projection of the I population to the E population . A modulatory input to the cells in the alpha frequency band can be interpreted as a slow variation of the input currents , hence appealing to the results shown in Fig 2 , we can expect the gamma power and frequency to vary with the alpha phase . Under these circumstances , the STH of the E population now displayed oscillations in the gamma as well as in the alpha band ( Fig 3A and 3B ) . These observations are also reflected in the power spectral density , which now has peaks in these two frequency bands ( Fig 3C ) . The introduction of the alpha rhythm not only decreased the gamma power , in accordance to experimental studies [34] , but also broadened the peak , reflecting the variation of gamma oscillation frequency with alpha phase . We applied a wavelet analysis to determine how the power of the gamma rhythm was coupled to the phase of the alpha rhythm . The alpha band activity was visible as a red-yellow horizontal band in the spectrogram centered around 10 Hz with only modest variations in power and frequency ( Fig 3D ) . The gamma power was modulated by alpha phase , which was reflected in a periodic sequence of transient yellow-green blobs ( Fig 3D , below ) . These blobs were locked to the peaks of the alpha cycle in the E population and the troughs in the I population ( Fig 3E ) . The phase of the alpha oscillations in the E and the I populations are approximately 180° apart , which indicates that high activity of the I population inhibits the E population . The inhibitory population thus serves to gate the activity of the excitatory neurons . If the drive from the pulvinar to the cortex projects predominantly to the inhibitory cells , this predicts that when excitatory neurons in the pulvinar fire , the excitatory neurons in the cortex are inhibited . This would mean that the alpha rhythm as measured in terms of excitatory neurons in cortex is fully out of phase relative to the activity in excitatory cells of the thalamus , given that axonal delays from pulvinar to cortex are significantly shorter than an alpha cycle . In order to quantify how alpha phase modulates communication we studied uni-directionally connected networks of two neuronal populations , each representing a cortical area as described in the METHODS section ( Fig 1 ) . We used gamma band coherence between these populations as index for the level of information transmission , communication for short . We varied both the relative alpha phase between the populations as well as alpha power . The following effects of alpha phase difference were obtained using a intermediate value of the alpha amplitude of 23 pA . The relative alpha phase difference had a strong effect on the coherence in the gamma band ( Fig 4A ) . The optimal alpha phase difference ( Δφ+ ) for communication , i . e . the one resulting in the highest gamma coherence was approximately -90° , while the least optimal alpha phase difference ( Δφ- ) , was approximately 90° . Maximal difference in gamma coherence between Δφ+ and Δφ- was obtained when the alpha oscillation was strong enough to reduce gamma power in the sending network to zero at the peak of the alpha phase ( i . e . when inhibitory activity was maximal ) . The alpha phase-difference leading to maximal gamma coherence corresponds to the state when area 1 leads area 2 by approximately 25 ms , or 1 gamma cycle . The most optimal and least optimal alpha phase for communication differ 180° . The overall firing rates of the excitatory and inhibitory populations in area 1 and 2 did not vary by more than 2% with alpha phase difference and can thus not be solely responsible for the much larger modulation of the gamma coherence ( Fig 4B ) . The coherence measure receives contributions from two different properties of a signal , an amplitude contribution , increasing when fluctuations across trials in amplitude between the two areas are similar , and a phase contribution , increasing when phases differences vary less across trials [35] . We were interested in determining on which component of the coherence the alpha modulation had the biggest effect . Fig 4C and 4D show the amplitude coherence and phase coherence , respectively . The amplitude coherence is very high , and strongly biased , as expected because the amplitudes in each area fluctuate similarly due to the common alpha modulation , even when compared across shuffled trials . However , the level of amplitude coherence does not vary strongly with alpha phase; rather the biggest effect is found for the gamma phase coherence . This indicates that the phases of the gamma oscillations synchronize better across areas when there is an optimal alpha phase difference . In Fig 4E the coherence spectrum for Δφ+ ( blue ) and Δφ- ( green ) is shown . The coherence in the alpha band is close to unity because of the synchronized alpha input and differs little between most and least optimal phase . There is however a clear difference in coherence around the gamma frequencies ( 30–50 Hz ) as well as a smaller harmonic effect around 70 Hz . These frequency bands match the peak locations in the power spectrum shown in Fig 3C . These results were obtained by modulating the input to the inhibitory neurons with an alpha oscillation . Similar results were obtained by modulating the input to the excitatory neurons ( S1 Fig ) , except that the alpha phases were shifted by approximately 180 degrees with respect to the thalamic rhythms . Next we modulated the alpha amplitude over a biologically relevant range to determine how this influenced the advantage of the optimal phase difference relative to least optimal phase difference . The rastergrams of neuronal firing under the lowest , middle and highest alpha amplitude can be found in S2 Fig . The modulation depth of the gamma coherence with phase difference increased with higher alpha amplitude ( Fig 4F ) . Higher alpha amplitudes increased gamma coherence for Δφ+ , and decreased it for Δφ- . The degree with which the alpha phase difference can increase gamma coherence had a sigmoidal dependence on alpha power . For lower values of the alpha amplitude no difference in gamma coherence existed , with further increases in amplitude the difference in gamma coherence increased strongly . The value of gamma coherence at the optimal alpha phase saturates with alpha power , hence further increases in alpha power do not increase maximum gamma coherence . The previous and following analyses were all performed for an intermediate level of alpha modulation amplitude equal to 23 pA ( see S2B Fig for rastergram ) , such that there was realistic gamma activity at all alpha phases . We implemented the alpha modulation as a sin2 function ( while dividing the frequency by 2 such that there was still an effective frequency of 10 Hz ) , which has a nonzero average , so that an increased modulation depth also increased the average input to the inhibitory neurons in cortex . We considered this the more biologically plausible setting [25 , 36]: as stronger alpha modulation of the cortex would entail more incoming spikes from the pulvinar . In order to test whether the modulation of gamma coherence with increasing alpha amplitude was not purely due to the increases or decreases of the firing rate , we did a control simulation with the alpha modulation implemented as a sine function , which has a zero mean . Cortical firing rates for this parameter setting did not vary with increasing alpha amplitude , however the modulation of gamma coherence nevertheless remained ( S3 Fig ) . In conclusion , gamma coherence , used here as an index for cortical communication , can be reliably modulated by the relative alpha phase between two cortical areas . From our previous results it is clear that alpha modulation of cortical areas influences the gamma rhythm in such a way that gamma coherence is increased or decreased . Especially the phase synchronization of the gamma rhythms in area 1 and area 2 are strongly influenced ( Fig 4E ) . To see what the effect of this increased synchronization is on the coordination of the spiking activity of individual neurons we compared the two extreme situations of the best alpha phase difference ( Δφ+ ) and the worst alpha phase difference ( Δφ- ) . A wavelet analysis was used to determine the phases of the gamma rhythms in area 1 and area 2 ( see METHODS ) . The spiking activity in area 1 and area 2 were both strongly modulated by their local gamma rhythms ( Fig 5A and 5B ) . Modulation of the spiking activity in the first area was almost the same for both conditions ( Δφ+ vs . Δφ- , blue vs . green respectively ) , whereas in the second area the spiking activity was slightly stronger modulated for Δφ+ ( Fig 5B ) . A larger difference is found when we considered how the spiking activity of area 1 aligned to the gamma rhythm in area 2 ( Fig 5C ) . For Δφ+ there is a much stronger locking of area 1 spiking activity to the area 2 gamma rhythm than for Δφ- . Note that in both conditions the phase difference between area 1 spiking activity and the gamma rhythm of area 2 is about 90° , so even the condition with low coherence between the rhythms there is a certain amount of phase locking between spikes and the rhythms . To see how the gamma rhythm influences the transfer of individual spikes in both conditions we calculated the probability that a spike in a neuron in area 1 is followed by a spike in a neuron in area 2 within a window of 1 to 4 ms , which matches the delays due to synaptic activation in our model ( we did not include the effects of axonal delays ) . Our neurons never fired more than once in a 4 ms period , hence we could represent each neuron-neuron pair as a binary value . The number of spikes in area 2 was normalised over the different gamma phase bins in area 1 by randomly removing spikes from bins until each bin contained the same number of spikes . This was necessary to ensure that an increased probability is not just the result of there being more spikes in a certain gamma phase bin . We observed a strong alignment of the spike emission probability in area 2 to the gamma rhythm of area 1 for the Δφ+ condition , and a somewhat weaker alignment for the Δφ- condition ( Fig 5D ) . Hence , gamma phases of area 1 corresponding to a high spiking activity also lead to a higher probability that a spike in area 1 was followed by a spike in area 2 . When we quantified the probability that a spike in area 1 is followed by a spike in area 2 in terms of alignment to the gamma rhythm in area 2 , we found a different effect ( Fig 5E ) . There is still alignment for the optimal phase difference , but the gamma phases in area 2 corresponding to high spiking activity do not lead to a higher probability that a spike in area 1 is followed by a spike in area 2 . Instead , for the best alpha phase difference the probability modulation agrees well with the results from Fig 5C: Gamma phases of area 2 for which more area 1 spikes occur , lead to a higher probability that a spike in area 1 is followed by a spike in area 2 . Together , these results suggest that the gamma rhythm in area 1 is responsible for increasing the effectiveness of spike transmission . Concentrating area 1 spiking activity in certain gamma phases increases area 1’s impact on area 2 . The strength of this effect is modulated by the alpha phase difference between the two areas . The best alpha phase difference , for which we also found high gamma coherence , leads to a stronger modulation of the effectiveness of spike transmission by the gamma rhythm in area 1 . An important feature of the process of selective attention is the ability to bias the processing of certain incoming stimuli over others , ensuring that the relevant stimuli have a bigger impact on downstream areas . We were interested whether the proposed mechanism of shifting alpha phases could also be used to prepare a transmission channel before the onset of a stimulus , in order to bias the processing of one stimulus over another by attention . We investigated whether shifting alpha phases can not only enhance gamma coherence and modulate spike effectiveness over a longer period of time as shown in the previous sections , but can also have an effect on short transient effects like the onset of a stimulus . Inspired by work that showed that the timing of a stimulus with respect to an ongoing cortical oscillation modulates stimulus processing [26 , 27] , we wondered whether similar effects could be found in our network and especially whether these effects also show up when a stimulus is communicated from one cortical area to another . To investigate this we needed to present a stimulus to the model network . We modeled the incoming stimulus as an increase in excitatory input to the first neuronal population . This increase in excitatory input caused our network to go from a state with low amplitude gamma oscillations driven by an interneuron gamma ( ING ) mechanism , which mainly relies only on interneuron activity for the generation of gamma oscillations , to high amplitude gamma oscillations driven by a PING mechanism ( Fig 6A ) . There was a strong initial non-linear effect on the overall firing rate of the population , after which the activity of the network returned to a state similar to the previous sections . We investigated the transient effect the stimulus had shortly ( 30 ms ) after stimulus onset on the overall firing rate of the neuronal populations . We quantified the response of the first population to the stimulus as the increase in firing rate during 30 ms after stimulus onset , for different moments of stimulation with respect to the phase of the alpha rhythm ( Fig 6B; solid line ) . Since the alpha phase also modulates the firing rate without stimulation , we also quantified this baseline ( Fig 6B; dashed line ) and took the difference as the effective response gain caused by the stimulus ( Fig 6C ) . A two-times-as-strong response was found on optimal alpha phases where inhibition due to the alpha modulation was weak , compared to least optimal alpha phases where inhibition was strong . This suggests that the brain could enhance or suppress the processing of a stimulus by shifting its alpha phase , but only if it can anticipate with sufficient temporal accuracy when a stimulus would occur . Although , as mentioned in the introduction , some evidence exists for alpha phase adjustments in anticipation of a stimulus [28] , the brain is also able to enhance or suppress the processing of unexpected stimuli . We hypothesize this might happen by setting up dynamical networks , or transmission channels , across which the signals are preferentially processed . To quantify this we considered the effect of the relative phase difference between two populations on the transmission of the stimulus-induced increases in firing rate of the presynaptic population . As in Fig 1 the first population was connected to the second . This allowed us to quantify communication as the increase in firing rate in the second population , transmitted by the first population in response to a stimulus . The average firing rate over the duration of strongest response was used , which was during 30 ms after stimulus onset . The response of the second population depended both on where in the alpha phase of the first population the stimulus was presented , as well as on the relative alpha phase difference between area 1 and area 2 ( Fig 6D ) . For the alpha phase giving the optimal response in area 1 the effect of the alpha phase difference with area 2 shows the same pattern as the gamma coherence ( Fig 6E and 6F ) . Averaged over all stimulus onsets at the different alpha phases , the response in area 2 shows the same effect ( S4A Fig ) . Having the optimal alpha phase difference between both populations can lead to a twice-as-strong response in the second population when compared to having the least optimal alpha phase difference . So not only the alpha phase with respect to the incoming stimulus matters , but also the relative alpha phase between communicating areas . This shows that shifting the relative alpha phase can bias the processing of a stimulus by increasing its impact on a downstream area . Moreover a stimulus is most effectively transmitted by having the optimal alpha phase difference between successive cortical areas , while having the wrong alpha phase difference can dampen the effect a stimulus response has on downstream areas . A question that remains is whether this effect of the relative alpha phase on the impact of stimuli enhances the information that is sent from one area to another . Discriminating between different stimuli is a task that human and animal subjects can perform and that often relies on the activity of cortical areas [37] and we used this task to quantify information transfer from area 1 to area 2 . We quantified the performance of the two cortical areas in discriminating two different stimuli entering at the optimal alpha phase of area 1 , but arriving at different alpha phases in area 2 due to ( top-down ) modulation of the alpha phase difference . Each stimulus was implemented as an increase in input to a different subpopulation of the first cortical area . One can think of this as each population having slightly different stimulus preferences . These two subpopulations in area 1 each projected selectively to the two corresponding subpopulations in the second cortical area , such that there was no overlap in the feedforward projections ( Fig 7A ) . To quantify the discriminating abilities of the second area , we trained a support vector machine ( SVM ) and used this to decode which stimulus was presented to the first area . The SVM was trained on the average firing rate of the neurons in each subpopulation of the second area during 30 ms after stimulus onset , where the strongest stimulus response was observed . Classification performance depended strongly on alpha phase difference ( S5A Fig ) . The joint probability of the actually presented stimulus and its SVM classification was used to calculate the mutual information as described in the METHODS section . Fig 7B shows how the mutual information between area 1 and area 2 depends on the relative alpha phase difference , as well as on the alpha phase of area 1 at stimulus onset . When we take the optimal alpha phase of area 1 ( Fig 6C ) , and change the alpha phase difference between both areas we see a strong modulation of the mutual information sent from area 1 to area 2 ( Fig 7C ) , similar to the effect of the alpha phase difference on coherence ( Fig 4A ) and stimulus response ( Fig 6F ) . Averaged over all possible stimulus onsets in the alpha phase of area 1 we find the same effect ( S5B Fig ) . The relative alpha phase thus not only increases the impact of a stimulus presented to area 1 on the response in area 2 , but also increases the ability of the second area to effectively discriminate different stimuli based on inputs it received from the first area . Furthermore , when we modulate the difficulty of the discrimination task by reducing the number of neurons that can be used for decoding , we see that this dependence on alpha phase difference is only present for difficult tasks involving a smaller number of neurons ( Fig 7D ) . When discriminating stimuli are easy the alpha phase difference doesn't matter as decoding is always performed with 100% accuracy . This supports the idea that this alpha phase dependent mechanism would only be useful for top-down processes that need to increase the performance of processing sensory information in demanding situations with limited neural resources . The preceding results were obtained with an unidirectional connection between area 1 and area 2 whose effectiveness could be manipulated through alpha phase shifting . Brain areas are often reciprocally coupled , which would need to be modeled by bidirectional connections . To study how feedback connections influence communication we incorporated feedback connections in our model ( Fig 7A ) . This means we cannot quantify communication solely in terms of coherence , but we also need to assess the direction of communication through analyses utilizing Granger causality . In the bidirectional model simulations we varied the relative alpha phase difference between both populations . The gamma coherence showed two peaks as a function of phase difference ( Fig 8B ) . We hypothesized that each peak corresponded to a different direction of communication . This hypothesis was tested using the conditional Granger causality , which was conditioned on the common alpha input that both areas received ( see METHODS ) . The relative alpha phase indeed determined the directionality of communication between the two populations ( Fig 8 ) . For negative alpha phase differences ( Fig 8C ) the directionality was from area 1 to area 2 , while for positive phase differences ( Fig 8D ) the directionality was from area 2 to area 1 . Causal communication was mainly in the gamma band . Without aforementioned conditioning on the common alpha input , causality in the alpha band dominated . For an alpha phase difference of 0 degrees , the causality was similar in both directions ( Fig 8E ) though closer inspection revealed a constantly switching directionality ( S6 Fig ) , leading on average to lower gamma coherence ( Fig 8A ) . The fluctuations in Granger causality occurred on timescales of between 100 and 400 ms and were probably caused by small variations in power and phase of the gamma oscillations .
This study assessed how communication between two cortical areas , indexed by gamma coherence , can be influenced by modulation of the phase of alpha oscillations generated in the thalamus . Our results show that the phase difference between alpha rhythms in each area was a powerful modulator of the effectiveness and direction of cortical communication . Increases in alpha power can either improve or impair cortical communication , depending on the phase difference between the alpha generators . A key component of the underlying mechanism was the coupling between alpha phase and gamma power , which is often reported in experiments [25] ( Fig 3D and 3E ) . Cortical communication was quantified in several ways to overcome the possible limitations that each individual measure might have . All measures demonstrated that communication was modulated with respect to the alpha phase difference between two areas . The variation in gamma band coherence was primarily caused by modulation of the alpha phase difference ( Fig 4A ) . Furthermore , we found that the alpha phase difference influences how well the arrival of individual spikes in the transmitting area is phase locked to the gamma rhythm of the receiving area ( Fig 5C ) , thereby increasing the probability that a spike in the transmitting area evokes a spike in the receiving area . Our simulations further show that the alpha phase difference between two areas can bias the transfer of an incoming stimulus . The response in the receiving area caused by the response to the stimulus in the transmitting area was strongly modulated by the alpha phase difference ( Fig 6E and 6F ) . To investigate whether an enhanced impact between the two areas could be utilized to increase bandwidth for stimulus information transfer , we quantified performance on a stimulus discrimination task using mutual information . These results also showed a strong modulation of the information transferred from the transmitting to the receiving area as a function of the alpha phase difference ( Fig 7C ) . In a network with bidirectional connections between two cortical areas we were able to control the direction of communication by shifting the relative alpha phase . Granger causality analysis showed that gamma band communication was strong from area 1 to area 2 and weak in the reverse direction , when area 1 was leading in alpha phase ( Fig 8C ) . The direction of gamma band GC was reversed when area 2 was leading in alpha phase instead ( Fig 8D ) . This suggests that our observation on the link between information processing and alpha phase difference for unidirectional projections extends to bidirectional connections . Coupling in the alpha band between thalamus and neocortical regions is well-established by animal recordings [18 , 19 , 38] . A number of recent studies report evidence for alpha-gamma cross-frequency coupling in human subjects that is consistent with the proposed model . Roux and coworkers used a beam-forming approach to show that the phase of alpha oscillations in virtual sensors located near the thalamus correlates with the amplitude of gamma oscillations in the early visual cortex [39] . A transfer entropy analysis of these data further showed that alpha phase affects gamma amplitude with a delay of about 16 ms . This is similar to physiological estimates of the transmission delay between the regions and follows from our model without explicitly building this in . Malekmohammadi and coworkers recorded simultaneously from cortex using an ECoG grid and from thalamus using depth electrodes . Their measurements show a coupling between the theta phase of oscillatory activity in the thalamus and the amplitude of beta band oscillations in cortex [40] ( see also [41] ) . Taken together , these studies demonstrate a coupling between slow-frequency oscillations in thalamus to gamma band activity in cortex in human subjects , thereby indicating that the model studied here may be more broadly applicable to the human visual system as well . The functional relevance of the phase of a low frequency oscillation ( alpha or theta band ) can be assessed by investigating how the detection probability of stimuli depends on the oscillatory phase at which they are presented . When visual stimuli were presented at a contrast close to detection threshold , the alpha phase for hits was different compared to that of the misses [42] . This was only true for phases calculated relative to oscillations in the theta and alpha band . Another experiment has demonstrated that not only is the detection probability modulated by alpha and theta oscillations but also that there is a phase difference in the alpha and theta frequency band between attended and unattended objects [43] . When the cyclical variation of detection probability is interpreted as being due to an oscillatory variation of neural excitability , it implies that there is a phase difference between the respective retinotopic areas responsible for processing the respective attended and unattended stimuli . This is consistent with the phase shifting between cortical areas hypothesized in the model . Recent experiments indicate that the thalamus , specifically the pulvinar , may be a major player in the coordination of information transmission by coordinating the synchronization between cortical areas . The simulation results reported here are relevant in relation to two recent experiments by Saalmann et al . [20] and Zhou et al . [44] . In Saalmann et al ( experiment 1 ) , monkeys are first cued to the relevant spatial location for the task , after which the stimulus is presented [20] . Upon stimulus presentation they have to respond according to the stimulus presented at the target location by either immediately releasing a lever or holding on and releasing it at a later time . Signals are recorded from the ventral pulvinar , V4 and area TEO during this task , and the analysis is conducted on responses after cue onset but before stimulus onset . The responses when attention is directed into RF are compared to when attention is directed to a location outside the RF . Within cortex the coupling between alpha phase and gamma power increases during attention . There is alpha band ( 10–15Hz ) coherence between the local field potentials ( LFPs ) in TEO and pulvinar , V4 and pulvinar as well as between TEO and V4 . This alpha-band coherence also increased with attention . In addition , there was a small increase in gamma coherence , but only for the TEO-V4 pairs . The direction of interaction between brain regions was investigated using Granger causality . When attention was directed into the receptive fields , the alpha-band Granger causality was increased in the direction from the pulvinar to cortex , but between the two cortical areas it was unaltered . The proposed interpretation of these results is that alpha oscillations in the pulvinar direct alpha oscillations in cortical areas that act to align gamma oscillations between areas , thereby improving information transmission [45] . Our model supports the potential role of alpha phase shifts in increasing efficiency of information transmission , reflected in enhanced gamma coherence and mutual information between firing rate fluctuations in the areas . As gamma coherence both depends on the degree of alpha alignment and whether sufficient activation is provided by bottom-up stimulus-related inputs or top-down inputs reflecting cognitive factors , the model would predict that when the stimuli are presented the intracortical gamma-band Granger causality would increase above its value that was measured during the cue period in [45] . Zhou and coworkers ( experiment 2 ) use a similar but not identical task and measured from ( ventral lateral ) pulvinar , V4 and TE ( a different area that is adjacent to TEO and has a different connectivity profile with the pulvinar [46] ) . In their task there is a cue in the center of the visual field that indicates the relevant location ( target ) in the stimulus array , which is comprised of different objects . The subject had to detect a small change in the target object and respond by making a saccade and the subject had to ignore changes in non-target objects . On some trials the cue preceded the onset of the stimulus array , thereby providing a measurement of how cortical networks re-organize to reflect the new focus of attention , prior to the onset of the stimulus response . In Figure S4 of the work by Zhou and coworkers the Granger causality between LFP signals was determined in the period prior to stimulus onset . The gamma band GC from V4 to TE was enhanced by attention , and it was weak and not modulated by attention the other way around . The pulvinar-V4 GC was frequency selective , with the gamma band dominating from V4 to pulvinar , which was also strongly modulated with attention , and lower frequency band dominating in the pulvinar to V4 GC . In contrast to the study by Saalmann and coworkers , the alpha band GC from pulvinar to cortex was not modulated by attention . Interestingly our model does not require an increase in Granger causality ( or coherence ) in the alpha band from the pulvinar to the cortex , since it’s only the phase of cortical alpha oscillations that is modulated . A possible interpretation of these results , which is consistent with our model , is that gamma band is representing stimulus-related information going from V4 to pulvinar , whereas the alpha band drive from pulvinar to V4 modulates the effectiveness of the V4-TE communication . Each band could reflect the activity of a different set of neurons , the balance between which varies with the particular group of neurons generating the LFP signal . The differences in findings of experiment 1 and 2 can thus be caused both by differences in task setup as well as recording sites . The findings of either of the studies do not contradict the results of our model , though they raise several issues that could be investigated with additional model simulations . Our model is a severe simplification of the pulvinar-cortical network . Here we describe several aspects that are important for improving our model in future work . A key simplification of our model is representing the pulvinar activity simply as an alpha oscillation . It would be important to understand how this alpha oscillation could be generated by a network of spiking neurons . Our mechanism requires the generation of alpha oscillations with different phases depending on attentional demand . The existence of such a mechanism is supported by the finding of phase diverse thalamic projections in the cat thalamus . In the lateral geniculate nucleus of the thalamus , high threshold thalamic cortical ( TC ) neurons ( HTC ) are a subset of TC neurons that produce bursts activated by a cholinergic projection or by activation of metabotropic glutamate receptors [47–49] . This leads to the situation that when the alpha oscillation is generated by the HTC , the remaining TC can fire at different phases . The same area can thus project alpha with two different phases relative to the local alpha oscillation . In vitro studies , controlling the level of depolarization , shows that a finer grained control of the phase is also possible [47 , 48] . Under the assumption that neurons in the appropriate location of the primate pulvinar have similar properties , this extensive set of experiments support the notion that pulvinar excitatory drives can be tuned in terms of alpha phase , which is a necessity for the proposed mechanism . Therefore , extending the model with a physiologically realistic network model of the pulvinar alpha generator would be a great addition . There is quite some evidence that the pulvinar-cortical interactions operate in both directions . The pulvinar is not only influencing cortex , but the cortex also projects back to the pulvinar [23] . Expanding our model by making the cortical areas influence the pulvinar would help to provide insight into the role of these feedback projections . Modeling stimulus representations in the pulvinar [24 , 50] and how bottom-up sensory information and top-down attentional control interact in the pulvinar could further increase our understanding of the pulvinar-cortical network . Finally , an important aspect that we have not implemented in our model is the laminar organization of the cortex . Different cortical laminae support the processing of information from respectively cortical and subcortical sources . Feedforward projections from the pulvinar project to layer 4 of the cortex , while feedback projections from the pulvinar project to more superficial cortical layers [51–53] . This would cause the pulvinar to influence the higher and lower areas in the visual hierarchy in mechanistically different ways , relying on different cell types and circuitry . In addition , cortical oscillations have a laminar profile , in terms of the origin of alpha oscillations [54 , 55] relative to faster oscillations [56] , which can differ going from earlier sensory areas ( V1 , A1 , S1 ) to the ones further up in the hierarchy ( V4 , IT ) [57] . Taking the laminar structure of the cortex into account is essential for making accurate predictions for electrophysiological measurements made with laminar electrical probes and should be considered in future work . Communication through coherence ( CTC ) was proposed as a principle to dynamically generate networks between cortical areas [2] . In its original form , it suggested that for areas to communicate effectively , they had to oscillate at the same frequency and the phase difference had to have a suitable value to align windows of excitability , for instance generated by a properly timed and synchronized inhibition [58] . The original principle of CTC did not come with a specific biophysical mechanism , but with predictions that were born out in experiment [59] . The lack of mechanism led to alternative interpretations [60 , 61] and modeling work arguing for it [62 , 63] and against it [64] . Experimentally , it was pointed out that gamma oscillations varied in frequency across time with a mean depending on the size and contrast of the stimulus [65] . This would make CTC difficult because the receiving area has to adjust to the ever-changing frequency . Nevertheless we now know that this is possible [66] , presumably due to the entrainment properties of PING circuits . In a recent review paper [6] , CTC has recently been revised to account for the new experimental results . In anesthetized monkeys , the relation between the probability of a spike in V1 eliciting a spike and V2 as a function of LFP phase was investigated [67] . CTC predicts that the spiking probability should be highest at the best local phase ( i . e . in V2 ) , but in experiment it was found to be at the best V1 phase , for which the V1 activity was the highest . Hence , the gamma was most effective at the time it corresponded to the highest number of active neurons . This is consistent with our model results ( Fig 5 ) . It supports a phase-based mechanism in the alpha band , which can influence information transmitted in the gamma band . Recently simulations have shown how a network with cross-frequency coupling can switch between a communication through coherence mechanism and a mechanism similar to the theta-gamma mechanism found in hippocampus [68] . This latter mechanism has been suggested to also play a role in visual processing [69] . Further investigations need to be conducted to find out how alpha phase changes would influence this mechanism . Our model is consistent with several recent experimental findings . To further validate our model we here formulate explicit predictions that can be tested by re-analyzing existing electrophysiological data or by conducting additional experiments . A clear prediction from our model is that the alpha phase difference between two connected cortical areas should be correlated with the gamma coherence between both areas . Such experiments can be conducted in animals using intracranial field recordings or in humans using MEG . A paradigm can be used in which the need for communication between the regions is manipulated; e . g . an attention type of paradigm . The alpha phase difference between two cortical areas should be different when comparing the attention into the receptive field condition to that with attention directed outside the receptive field condition . On a trial by trial basis , the difference in alpha phase should be correlated with gamma coherence , as well as task performance . Finally , methods such as DREADD [70] and optogenetics [71] can be used to actively manipulate the phase of cortical alpha oscillations in animal studies . These techniques are well suited to target very specific groups of neurons in the cortex or pulvinar . By active individual manipulation of the alpha phase in two connected cortical areas we predict an influence on gamma coherence between these areas . These techniques are currently well-established for use in rodents , but are in the process of being adapted for primates use [72] . We have shown a possible role for the alpha rhythm in coordinating cortical communication . By controlling the alpha phase difference between cortical areas we were able to influence the effectiveness of communication and the processing of stimuli . In bidirectionally coupled networks the alpha phase difference determined the direction of communication . The pulvinar would be an ideal candidate for controlling this mechanism . We formulated a number of experimental tests to support the hypothesized mechanisms and further clarify the role of the pulvinar in the process of selective attention .
To study communication between neuronal populations in different cortical areas we modeled each population as a local network of strongly interconnected neurons ( Fig 2 ) . Each population consisted of 400 regular spiking ( RS ) excitatory pyramidal neurons , 75 fast spiking ( FS ) interneurons and 25 low threshold spiking ( LTS ) interneurons [73] . Connectivity within each population was all-to-all: each neuron projected to all other neurons , representing the strong connectivity of a local cortical population . In most simulations , two of these populations were connected , each population representing a part of a different cortical area within the visual cortex . Depending on the simulation experiment , the populations were connected either uni-directionally or bi-directionally . Each neuron in each population received an independent random noise current inputs that was uncorrelated between neurons . Unless stated otherwise , the input was modulated with a 10 Hz ( alpha frequency ) sinusoidal oscillation , simulating the coordinated input activity from the pulvinar . In order to investigate cortical communication the following analyses were performed . All simulations and analyses were performed in MATLAB R2012b . Our neuron model was numerically integrated using an adaptation of Euler’s method [31] . Every simulation step consisted of 2 sequential steps of 0 . 5 ms for numerical stability . After each simulation step spikes were detected . Specifically , we defined spike times as the time at which the spike reached its peak potential of 30 mV . The mean firing rate was determined separately for each neuron type as the number of firings of a neuron divided by the trial time , averaged over all neurons of the same type r=1NT∑ini with r being the firing rate , N the number of neurons of a certain type , T the trial duration and ni the number of firings of neuron i . We defined rRS , rFS and rLTS to be the firing rates for the regular spiking , fast spiking and low threshold spiking neurons , respectively . Many of the analyses were performed on the spike timing histograms ( STHs ) for each neuron type seperately . The STHs were defined as the total number of firings of a neuron type within a certain time bin STH=∑iXi ( t ) Where Xi ( t ) ={1 , iffji∈ ( t , t+Δt ) 0otherwise with Xi ( t ) determining whether neuron i fired within the time bin at time t , fij the time of the jth spike of neuron i . The time bin had a width of Δt = 1ms . | Cortical oscillations have been linked to the process of communication between two brain areas . Here we investigated how a third area could control communication between two other brain areas . We find that the phase of a slower alpha-band oscillation is able to influence the power of faster gamma oscillations . By changing phase differences between the slower oscillation in two areas , a third area is able to control the amount of information flow . In a network with bi-directional connections , the direction of communication is also controlled by this phase difference . Our results suggest that the pulvinar could coordinate communication between different brain areas . This area could have a central role in prioritizing the processing of sensory information that is most relevant for the task at hand . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Models",
"&",
"methods"
] | [
"cognitive",
"science",
"action",
"potentials",
"medicine",
"and",
"health",
"sciences",
"membrane",
"potential",
"brain",
"social",
"sciences",
"electrophysiology",
"neuroscience",
"cognitive",
"psychology",
"gamma",
"spectrometry",
"extraction",
"techniques",
"crystallogr... | 2017 | Top-down control of cortical gamma-band communication via pulvinar induced phase shifts in the alpha rhythm |
Recent advances in next-generation sequencing and computational technologies have enabled routine analysis of large-scale single-cell ribonucleic acid sequencing ( scRNA-seq ) data . However , scRNA-seq technologies have suffered from several technical challenges , including low mean expression levels in most genes and higher frequencies of missing data than bulk population sequencing technologies . Identifying functional gene sets and their regulatory networks that link specific cell types to human diseases and therapeutics from scRNA-seq profiles are daunting tasks . In this study , we developed a Component Overlapping Attribute Clustering ( COAC ) algorithm to perform the localized ( cell subpopulation ) gene co-expression network analysis from large-scale scRNA-seq profiles . Gene subnetworks that represent specific gene co-expression patterns are inferred from the components of a decomposed matrix of scRNA-seq profiles . We showed that single-cell gene subnetworks identified by COAC from multiple time points within cell phases can be used for cell type identification with high accuracy ( 83% ) . In addition , COAC-inferred subnetworks from melanoma patients’ scRNA-seq profiles are highly correlated with survival rate from The Cancer Genome Atlas ( TCGA ) . Moreover , the localized gene subnetworks identified by COAC from individual patients’ scRNA-seq data can be used as pharmacogenomics biomarkers to predict drug responses ( The area under the receiver operating characteristic curves ranges from 0 . 728 to 0 . 783 ) in cancer cell lines from the Genomics of Drug Sensitivity in Cancer ( GDSC ) database . In summary , COAC offers a powerful tool to identify potential network-based diagnostic and pharmacogenomics biomarkers from large-scale scRNA-seq profiles . COAC is freely available at https://github . com/ChengF-Lab/COAC .
Single cell ribonucleic acid sequencing ( scRNA-seq ) offers advantages for characterization of cell types and cell-cell heterogeneities by accounting for dynamic gene expression of each cell across biomedical disciplines , such as immunology and cancer research [1 , 2] . Recent rapid technological advances have expanded considerably the single cell analysis community , such as The Human Cell Atlas ( THCA ) [3] . The single cell sequencing technology offers high-resolution cell-specific gene expression for potentially unraveling of the mechanism of individual cells . The THCA project aims to describe each human cell by the expression level of approximately 20 , 000 human protein-coding genes; however , the representation of each cell is high dimensional , and the human body has trillions of cells . Furthermore , scRNA-seq technologies have suffered from several limitations , including low mean expression levels in most genes and higher frequencies of missing data than bulk sequencing technology [4] . Development of novel computational technologies for routine analysis of scRNA-seq data are urgently needed for advancing precision medicine [5] . Inferring gene-gene relationships ( e . g . , regulatory networks ) from large-scale scRNA-seq profiles is limited . Traditional approaches to gene co-expression network analysis are not suitable for scRNA-seq data due to a high degree of cell-cell variabilities . For example , LEAP ( Lag-based Expression Association for Pseudotime-series ) is an R package for constructing gene co-expression networks using different time points at the single cell level [6] . The Partial information decomposition ( PID ) algorithm aims to predict gene-gene regulatory relationships [7] . Although these computational approaches are designed to infer gene co-expression networks from scRNA-seq data , they suffer from low resolution at the single-cell or single-gene levels . In this study , we introduced a network-based approach , termed Component Overlapping Attribute Clustering ( COAC ) , to infer novel gene-gene subnetwork in individual components ( the subset of whole components ) representing multiple cell types and cell phases of scRNA-seq data . Each gene co-expression subnetwork represents the co-expressed relationship occurring in certain cells . The scoring function identifies co-expression networks by quantifying uncoordinated gene expression changes across the population of single cells . We showed that gene subnetworks identified by COAC from scRNA-seq profiles were highly correlated with the survival rate of melanoma patients and drug responses in cancer cell lines , indicating a potential pathobiological application of COAC . If broadly applied , COAC can offer a powerful tool for identifying gene-gene networks from large-scale scRNA-seq profiles in multiple diseases in the on-going development of precision medicine .
In this study , we present a novel algorithm for inferring gene-gene networks from scRNA-seq data . Specifically , a gene-gene network represents the co-expression relationship of certain components ( genes ) , which indicates the localized ( cell subpopulation ) co-expression from large-scale scRNA-seq profiles ( Fig 1 ) . Specifically , each gene subnetwork is represented by one or multiple feature vectors , which are learned from the scRNA-seq profile of the training set . For the test set , each gene expression profile can be transformed to a feature value by one or several feature vectors which measure the degree of coordination of gene co-expression . Since the feature vectors are learned from the relative expression of each gene , batch effects can be eliminated by normalization of relatively co-expressed genes ( see Methods ) . In addition to showing that COAC can be used for batch effect elimination , we further validated COAC by illustrating three potential pathobiological applications: ( 1 ) cell type identification in two large-scale human scRNA-seq datasets ( 43 , 099 and 43 , 745 cells respectively , see Methods ) ; ( 2 ) gene subnetworks identified from melanoma patients-derived scRNA-seq data showing high correlation with survival of melanoma patients from The Cancer Genome Atlas ( TCGA ) ; ( 3 ) gene subnetworks identified from scRNA-seq profiles which can be used to predict drug sensitivity/resistance in cancer cell lines . We collected scRNA-seq data generated from 10x scRNA-seq protocol [7 , 8] . In total , 14 , 032 cells extracted from peripheral blood mononuclear cells ( PBMC ) in systemic lupus erythematosus ( SLE ) patients were used as the case group and 29 , 067 cells were used as the control group ( see Methods ) . For the case group , we used 12 , 277 cells for the training set and the remaining 1 , 755 cells for the validation set . For the control group , we used 25 , 433 cells for the training set and 3 , 634 for the validation set . After filtering with average correlation and average component ratio thresholds ( see Methods ) , we obtained 93 , 951 co-expression subnetworks ( gene clusters with components ) by COAC . We transformed these co-expression gene clusters to feature vectors . Features whose variance distribution was significantly different in the case group versus the control group were kept ( see Methods ) . Using a t-SNE algorithm implemented in the R package-tsne [9] , we found that the single cells ( from the case group ) which were retrieved directly from the patients can be more robustly separated from the control group cells ( Fig 2B ) , comparing to the original data ( Fig 2A ) without applying COAC . Thus , the t-SNE analysis reveals that batch effects can be significantly reduced by COAC ( Fig 2 ) . We next turned to examine whether COAC can be used for cell type identification . We collected a scRNA-seq dataset of 14 , 448 single cells in an IFN-β stimulated group and 14 , 621 single cells in the control group [8] . To remove factors caused by the stimulation conditions or experimental batch effects , we selected 13 , 003 cells in the IFN-β stimulated group and 13 , 158 cells in the control group as the training set to obtain homogeneous feature vectors for each cell . The remaining scRNA-seq data are used as the validation set . We generated the gene subnetworks by COAC and transformed the subnetworks into feature vectors for individual cells ( see Methods ) . We found that cells from IFN-β stimulated and control groups were separated significantly ( Fig 3A ) by t-SNE [9] . However , without applying COAC cells from the IFN-β stimulated and control groups are uniformly distributed in the whole space ( Fig 3B ) , suggesting that components which separate IFN-β stimulated cells from control cells were eliminated from the feature vector identified by COAC . We further collected a scRNA-seq dataset including a total of 43 , 745 cells with well-defined cell types from a previous study [10] . We built a training set ( 21 , 873 cells ) and a validation set ( 21 , 872 cells ) with approximately equivalent size . In the training set , we generated co-expression subnetworks as the feature vector by COAC . For the validation set , we grouped the total cells into five main categories as described previously [10] . Fig 3C shows that COAC-inferred subnetworks can be used to distinguish five different cell types with high accuracy ( cell types for 83 . 05% cells have been identified correctly ) in the t-SNE analysis , indicating that COAC can identify cell types from heterogeneous scRNA-seq profiles . We next inspected potential pathobiological applications of COAC in identifying possible prognostic biomarkers or pharmacogenomics biomarkers in cancer . We next turned to inspect whether COAC-inferred gene co-expression subnetworks can be used as potential prognostic biomarkers in clinical samples . We identified gene subnetworks from scRNA-seq data of melanoma patients [11] . Using a feature selection pipeline , we filtered the original subnetworks according to the difference of means and variances between two different groups ( e . g . , malignant cells versus control cells ) to prioritize top gene co-expression subnetworks ( S1A Fig ) . We collected the bulk gene expression data and clinical data for 458 melanoma patients from the TCGA website [12] . Applying COAC , we identified two gene co-expression subnetworks with the highest co-expression correlation in malignant cells compared to control cells ( S1B Fig ) . For each subnetwork , we then calculated the co-expression correlation in bulk RNA-seq profiles of melanoma patients . Using the rank of co-expression values of melanoma patients , the top 32 patients were selected as group 1 and the tail 32 patients were selected as group 2 . Log rank test was employed to compare the survival rate of two groups [13] . We found that gene subnetworks identified by COAC from melanoma patients-derived scRNA-seq data can predict patient survival rate ( Fig 4A and Fig 4B ) . KRAS , is an oncogene in multiple cancer types [14] , including menaloma [15] . Herein we found a co-expression among KRAS , HADHB , and PSTPIP1 , can predict significantly patient survival rate ( P-value = 4 . 09×10−5 , log rank test , Fig 4B ) . Thus , regulation of KRAS-HADHB-PSTPIP1 may offer new a pathobiological pathway and potential biomarkers for predicting patient’s survival in menaloma . We next focused on gene co-expression subnetworks in several known melanoma-related pathways , such as the MAPK , cell-cycle , DNA damage response , and cell death pathways [16] by comparing the differences in means and variances between T cell and other cells using COAC ( see Methods ) . For each gene co-expression subnetwork identified by COAC , we selected 32 patients who had enriched co-expression correlation and 32 patients who had lost a co-expression pattern . We found that multiple COAC-inferred gene subnetworks predicted significantly menaloma patient survival rate ( Fig 4C–4F ) . For example , we found that BRAF-PSMB3-SNRPD2 predict significant survival ( P-value = 0 . 0058 , log rank test . Fig 4C ) , revealing new potential disease pathways for BRAF melanoma . CDKN2A , encoding cyclin-dependent kinase Inhibitor 2A , plays important roles in melanoma [17] . Herein we found a potential regulatory subnetwork , RBM6-CDKN2A-MRPL10-MARCKSL , which is highly correlated with melanoma patients’ survival rate ( P-value = 0 . 019 , log rank test . Fig 4F ) . We identified several new potential regulatory subnetworks for TP53 as well , which is highly correlated with patient's survival rate as well ( Fig 4D and 4E ) . Multiple novel COAC-inferred gene co-expression subnetworks that are significantly associated with patient’s survival rate are provided in S2 Fig . Altogether , gene regulatory subnetworks identified by COAC can shed light on new disease mechanisms uncovering possible functional consequences of known melanoma genes and offer potential prognostic biomarkers in melanoma . COAC-inferred prognostic subnetworks should be further validated in multiple independent cohorts before clinical application . To examine the potential pharmacogenomics application of COAC , we collected robust multi-array ( RMA ) gene expression profiles and drug response data ( IC50 [The half maximal inhibitory concentration] ) across 1 , 065 cell lines from the Genomics of Drug Sensitivity in Cancer ( GDSC ) database [18] . We selected six drugs in this study based on two criteria: ( i ) the highest variances of IC50 among over 1 , 000 cell lines , and ( ii ) drug targets across diverse pathways: SNX-2112 ( a selective Hsp90 inhibitor ) , BX-912 ( a PDK1 inhibitor ) , Bleomycin ( induction of DNA strand breaks ) , PHA-793887 ( a pan-CDK inhibitor ) , PI-103 ( a PI3K and mTOR inhibitor ) , and WZ3105 ( also named GSK-2126458 and Omipalisib , a PI3K inhibitor ) . We first identified gene co-expression subnetworks from melanoma patients’ scRNA-seq data [11] by COAC . The COAC-inferred subnetworks with RMA gene expression profiles of bulk cancer cell lines were then transformed to a matrix: each column of this matrix represents a feature vector and each row represents a cancer cell line from the GDSC database [18] . We then trained an SVM regression model using the LIBSVM [19] R package with default parameters and linear kernel ( see Methods ) . We defined cell lines whose IC50 were higher than 10 μM as drug-resistant cell lines ( or non-antitumor effects ) , and the rest as drug sensitive cell lines ( or potential antitumor effects ) . As shown in Fig 5A–5F , the area under the receiver operating characteristic curves ( AUC ) ranges from 0 . 728 to 0 . 783 across 6 drugs during 10-fold cross-validation , revealing high accuracy for prediction of drug responses by COAC-inferred gene subnetworks . To illustrate the underlying drug resistance mechanisms , we showed two subnetworks identified by COAC for SNX-2112 ( Fig 5G ) and BX-912 ( Fig 5H ) respectively . SNX-2112 , a selective Hsp90 ( encoded by HSP90B1 ) inhibitors , has been reported to have potential antitumor effects in preclinical studies , including melanoma [20 , 21] . We found that several HSP90B1 co-expressed genes ( such as CDC123 , LPXN , and GPX1 ) in scRNA-seq data may be involved in SNX-2112’s resistance pathways ( Fig 5G ) . GPX1 [22] and LPXN [23] have been reported to play crucial roles in multiple cancer types , including melanoma . BX-912 , a PDK1 inhibitor , has been shown to suppress tumor growth in vitro and in vivo [24] . Fig 5H shows that several PDK1 co-expressed genes ( such as TEX264 , NCOA5 , ANP32B , and RWDD3 ) may mediate the underlying mechanisms of BX-912’s responses in cancer cells . NCOA5 [25] and ANP32B [26] were reported previously in various cancer types . Collectively , COAC-inferred gene co-expression subnetworks from individual patients’ scRNA-seq data offer the potential underlying mechanisms and new biomarkers for assessment of drug responses in cancer cells .
In this study , we proposed a network-based approach to infer gene-gene relationships from large-scale scRNA-seq data . Specifically , COAC identified novel gene-gene co-expression in individual certain components ( the subset of whole components ) representing multiple cell types and cell phases , which can overcome a high degree of cell-cell variabilities from scRNA-seq data . We found that COAC reduced batch effects ( Fig 2 ) and identified specific cell types with high accuracy ( 83% , Fig 3C ) in two large-scale human scRNA-seq datasets . More importantly , we showed that gene co-expression subnetworks identified by COAC from scRNA-seq data were highly corrected with patients’ survival rate from TCGA data and drug responses in cancer cell lines . In summary , COAC offers a powerful computational tool for identification of gene-gene regulatory networks from scRNA-seq data , suggesting potential applications for the development of precision medicine . There are several improvements in COAC compared to traditional gene co-expression network analysis approaches from RNA-seq data of bulk populations . Gene co-expression subnetwork identification by COAC is nearly unsupervised , and only a few parameters need to be determined . Since gene overlap among co-expression subnetworks is allowed , the number of co-expression subnetworks has a higher order of magnitude than the number of genes . Gene co-expression subnetworks identified by COAC can capture the underlying information of cell states or cell types . In addition , gene subnetworks identified by COAC shed light on underlying disease pathways ( Fig 4 ) and offer potential pharmacogenomics biomarkers with well-defined molecular mechanisms ( Fig 5 ) . We acknowledged several potential limitations in the current study . First , the number of predicted gene co-expression subnetworks is huge . It remains a daunting task to select a few biologically relevant subnetworks from a large number of COAC-predicted gene subnetworks . Second , as COAC is a gene co-expression network analysis approach , subnetworks identified by COAC are not entirely independent . Thus , the features used for computing similarities among cells are not strictly orthogonal . In the future , we may improve the accuracy of COAC by integrating the human protein-protein interactome networks and additional , already known , gene-gene networks , such as pathway information [27–29] . In addition , we could improve COAC further by applying deep learning approaches [30] for large-scale scRNA-seq data analysis . In summary , we reported a novel network-based tool , COAC , for gene-gene network identification from large-scale scRNA-seq data . COAC identifies accurately the cell types and offers potential diagnostic and pharmacogenomic biomarkers in cancer . If broadly applied , COAC would offer a powerful tool for identifying gene-gene regulatory networks from scRNA-seq data in immunology and human diseases in the development of precision medicine .
In COAC , a subnetwork is represented by the eigenvectors of its adjacency correlation matrix . In practice , the gene regulatory relationships represented by each subnetwork are not always unique . Those that occur in each subnetwork represent a superposition of two or several regulatory relationships , where each has a weight in gene subnetworks shown in S3A Fig . We thereby used multi-components ( i . e . , top eigenvectors with large eigenvalues ) to represent the co-expression subnetworks . As shown in S3B Fig , a regulatory relationship between two genes can be captured in different co-expression subnetworks . Herein , we integrated matrix factorization [31] into the workflow of closed frequent pattern mining [32] . Specifically , the set of closed frequent patterns contains the complete itemset information regarding these corresponding frequent patterns [32] . Here , closed frequent pattern is defined that if two item sets appear in the same samples , only the super one is kept . For a general gene expression matrix , to obtain a sparse distribution of genes in each latent variable , a matrix factorization method such as sparse principal component analysis ( PCA ) [33] can be chosen . In this study , because the scRNA-seq data matrix is highly sparse , singular value decomposition ( SVD ) is chosen for matrix factorization ( i . e . , the SVD of A is given by UσV* ) . The robust rank r is defined in the S1 Text . Components that are greater than rank r are selected and then each attribute is treated as the linearly weighted sum of components ( Di = wi1 P1 + wi2 P2 + wi3 P3 …wir Pr ) . The projection of gene distribution i over principal component j can be expressed as DitPj‖Di‖‖Pj‖ , where ‖Pj‖ = 1 . Then , D ( i , j ) =DitPj‖Di‖‖Pj‖=DitPj‖Di‖=wij‖Di‖ and −1<DitPj‖Di‖<1 . The projection of each attribute distribution over each principal component distribution is illustrated in S4A Fig . In practice , single cell data are always sparse . For component j , most elements in the collection of D ( i , j ) |j are zero . Several thresholds are determined by F-distribution . For a component j , the mean and the variance of collection D ( i , j ) |j is m and s2 . Then the F-distribution with degree of freedom 1 , and degree of freedom N-1 ( N is the number of attributes ) is: F ( 1 , N−1 ) ( x ) = ( x−m ) 2s2 ( 1 ) The P-value for a element x in collection D ( i , j ) |j is the extreme upper tail probability of this F-distribution . The threshold of the collection D ( i , j ) |j is divided into two groups . In one group , the P-value of all element should be below a pre-defined threshold . The detailed process for obtaining the thresholds is described in the S1 Text . Herein , the cutoff of P-value for F-distribution ranges from 0 . 01 to 0 . 05 . Subsequently , we defined the mapping rule using these thresholds . {1ifthresholdPj<DxtPj‖Dx‖<1 ( Gain ) 0ifthresholdNj<DxtPj‖Dx‖<thresholdPj ( Non−effect ) −1if−1<DxtPj‖Dx‖<thresholdNj ( Loss ) ( 2 ) The pipeline is shown in S4B and S4C Fig . In the ( 1/0 ) sparse matrix , each row represents a component while each column represents an attribute ( gene ) . The association rule is consisted of: ( i ) one is an attribute ( gene ) collection and ( ii ) the other is a component collection . The position in the binary distribution matrix of any pair with the Cartesian product of the two collections is always 1 . This position is shown in S4D and S4E Fig . For each association rule , the attribute collection should have maximal component collection . For example , for association rules {X Y Z} {M} , {X Y} {M} , {X Y} {M N} , only the maximal {X Y} {M N} is allowed . And the closed association rule states that if two rules have the same component collections , only the maximal attribute collection is preserved and kept . For association rules {X Y Z} {M N} , {X Y} {M N} , {Y Z} {M N} , and {X Z} {M N} , with the same component collection {M , N} , only the maximal {X Y Z} {M N} is kept , whereas the others are removed . The process of efficient enumeration of all significant association rules ( gene subnetwork ) is described in the S1 Text . The subnetwork and gene distribution of selected components are obtained directly by applying the association rule , and the gene subnetwork is treated as the largest connected component ( graph ) from co-expression networks of scRNA-seq profiles . Finally , two metrics are introduced for filtering . The average correlation among genes in each subnetwork is a measure of the homogeneity of genes with selected components . The average component ratio denotes the average of how much of the whole component space is occupied by the selected components . AverageCorrelation= ( 1n ( n−1 ) ) ∑i , j∈{X , Y , Z} , j , i≠jCorrelation ( Ai , Aj ) |M , N ( 3 ) ComponentRatioofAi=‖Ai‖2|selectedcomponents‖Ai‖2 ( 4 ) AverageComponentRatio=1N∑ComponentRatioofAi ( Ai ∈ attribute collection of a closed associate rule ) ( 5 ) The processes of obtaining the average correlation and the average component ratio are provided in the S1 Text . The final largest connected component subnetwork is represented by several eigenvectors with large eigenvalues , which are calculated from the correlation matrix . These eigenvectors are used to map each record of the gene expression profile into individual numerical values ( feature vectors ) . Featurevector=SFt/‖S‖2 ( ‖F‖2=1 ) ( 6 ) Where S is the gene expression vector for each cell , and F is the first eigenvector of the component matrix . If several principal components exist , then the feature value becomes the sum of components multiplied by the attenuation coefficient . Featurevector=SF1t/‖S‖2+ ( σ2/σ1 ) SF2t/‖S‖2+ ( σ3/σ1 ) SF3t/‖S‖2… ( ‖F1t‖2=1 , ‖F2t‖2=1… ) ( 7 ) Where σ1 , σ2 , σ3 , … , σv are the eigenvalues of the gene clustering ( subnetwork ) correlation matrix , and F1t , F2 , t… are the eigenvectors of gene clustering correlation matrix . The purpose of cell type alignment was to label cell types of each cell under different conditions . Cell types with the same labels under each condition were then clustered . Subsequently , differential expression analyses were performed for various conditions of each cell type . Finally , surrogate variable analyses [34] were performed to remove the batch effects . We used the limma [35] method ( S5B Fig ) for the differential expression analysis of the differently conditioned cell types . The scRNA-seq data ( GEO accession ID: GSE96583 ) that was used to test the batch effect elimination was collected from PBMC peripheral blood mononuclear cells of SLE patients [7 , 8] . In total , 14 , 032 cells with 13 aligned PBMC subpopulations under resting and interferon β ( IFN-β ) -stimulated conditions were collected [8] . In addition , we also collected 29 , 067 cells from two controls as the control group [7] . For the training dataset , the variances of the feature vectors ( COAC-identified subnetworks ) between the case group and the control group were calculated and was regarded as differential variances . The variances of the feature vectors of the merged group of the case group and the control group were regarded as background variances . For each feature , the ratio of the differential variance and background variance was defined as F-score , which measured how much this feature can distinguish cells in a case group versus a control group . The F-score distribution for 93 , 951 features is described in S6 Fig . Using a critical point of 2 . 4 as a threshold ( S6 Fig ) , 8 , 331 features with F-score higher than the threshold were kept . For comparison , we used 2 , 657 genes which were used as biomarkers previously as the feature vector [8] . The scRNA-seq data of mouse kidney with well-annotated cell types were collected from a previous study [10] . By stringent quality controls described previously [10] , a total of 43 , 745 cells selected from the original 57 , 979 cells were used in this study . The entire dataset was randomly divided into the training set ( 21 , 873 cells ) and the test set ( 21 , 872 cells ) . The detail of prediction model construction can be found in cell type alignment pipeline ( S5 Fig ) . For the validation part , cell type was predicted using the training model . For each cell , the scores for cell types were calculated . Then all cells were plotted by t-SNE algorithm [9] . The results of cell type prediction were displayed in the confusion matrix . We collected the melanoma patients’ scRNA-seq data with well-annotated cell types from a previous study [11] . The bulk RNA-seq data and clinical profiles for melanoma patients were collected from the TCGA website [13] . The gene expression values in the scRNA-seq dataset were transformed as log ( TPMij+1 ) , where TPMij refers to transcript-per-million ( TPM ) of gene i in cell j . The gene expression value in the bulk RNA-seq dataset was transformed in the same way . The sub-network list was obtained from melanoma scRNA-seq dataset [11] by COAC . Sub-networks then were transformed to feature vectors . Two top sub-networks with the highest co-expressed correlation in melanoma cell type and one top sub-network with the highest co-expressed correlation in T cells were evaluated . The co-expression values were calculated with RNA-seq gene expression of melanoma patients from TCGA [13] . Survival analysis was conducted using an R survival package [36] . We downloaded drug response data ( defined by IC50 value ) and gene bulk expression profiles in cancer cell lines from the GDSC database [18] . The component co-expression sub-networks were identified from the melanoma patients’ scRNA-seq data with well-annotated cell types from a previous study [11] . For scRNA-seq data , genes that had a ratio of expressed cells less than 0 . 03 were removed . Herein , we kept the top 0 . 1~0 . 01 percent subnetworks with the highest correlation as feature vectors . We predicted each drug’ IC50 value by LIBSVM [19] R package with default parameters and linear kernel . The ROC curves for the result of drug response were plotted using the R package . | Single-cell RNA sequencing ( scRNA-seq ) can reveal complex and rare cell populations , uncover gene regulatory relationships , track the trajectories of distinct cell lineages in development , and identify cell-cell variabilities in human diseases and therapeutics . Although experimental methods for scRNA-seq are increasingly accessible , computational approaches to infer gene regulatory networks from raw data remain limited . From a single-cell perspective , the stochastic features of a single cell must be properly embedded into gene regulatory networks . However , it is difficult to identify technical noise ( e . g . , low mean expression levels and missing data ) and cell-cell variabilities remain poorly understood . In this study , we introduced a network-based approach , termed Component Overlapping Attribute Clustering ( COAC ) , to infer novel gene-gene subnetworks in individual components ( subsets of whole components ) representing multiple cell types and phases of scRNA-seq data . We showed that COAC can reduce batch effects and identify specific cell types in two large-scale human scRNA-seq datasets . Importantly , we demonstrated that gene subnetworks identified by COAC from scRNA-seq profiles highly correlated with patients's survival and drug responses in cancer , offering a novel computational tool for advancing precision medicine . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods",
"and",
"materials"
] | [
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"clinical",
"research",
"design",
"engineering",
"and",
"technology",
"statistics",
"gene",
"regulation",
"computational",
"biology",
"cancers",
"and",
"neoplasms",
"biomarkers",
"oncology",
"research",
"design",
... | 2019 | A component overlapping attribute clustering (COAC) algorithm for single-cell RNA sequencing data analysis and potential pathobiological implications |
Rural populations in the Gran Chaco region have large prevalence rates of Trypanosoma cruzi infection and very limited access to diagnosis and treatment . We implemented an innovative strategy to bridge these gaps in 13 rural villages of Pampa del Indio held under sustained vector surveillance and control . The non-randomized treatment program included participatory workshops , capacity strengthening of local health personnel , serodiagnosis , qualitative and quantitative PCRs , a 60-day treatment course with benznidazole and follow-up . Parents and healthcare agents were instructed on drug administration and early detection and notification of adverse drug-related reactions ( ADR ) . Healthcare agents monitored medication adherence and ADRs at village level . The seroprevalence of T . cruzi infection was 24 . 1% among 395 residents up to 18 years of age examined . Serodiagnostic ( 70% ) and treatment coverage ( 82% ) largely exceeded local historical levels . Sixty-six ( 85% ) of 78 eligible patients completed treatment with 97% medication adherence . ADRs occurred in 32% of patients , but most were mild and manageable . Four patients showing severe or moderate ADRs required treatment withdrawal . T . cruzi DNA was detected by qPCR in 47 ( 76% ) patients before treatment , and persistently occurred in only one patient over 20–180 days posttreatment . Our results demonstrate that diagnosis and treatment of T . cruzi infection in remote , impoverished rural areas can be effectively addressed through strengthened primary healthcare attention and broad social participation with adequate external support . This strategy secured high treatment coverage and adherence; effectively managed ADRs , and provided early evidence of positive therapeutic responses .
Chagas disease ranks among the main neglected tropical diseases ( NTDs ) in Latin America and the Caribbean [1] . Trypanosoma cruzi , its etiologic agent , induces heart and digestive disease and reduces life expectancy in approximately 30–40% of the infected people [2 , 3] . The parasite infects 6–9 million people , the majority of which primarily were rural residents living in poverty with little access to healthcare services [4] . A well-known hotspot of Chagas disease and other NTDs is the Gran Chaco ecoregion which mainly extends over sections of Argentina , Bolivia , and Paraguay [5] . In rural villages across this region where Triatoma infestans is the only domestic vector , the seroprevalence of human T . cruzi infection in children younger than 15 years of age frequently ranged between 20% and 50% [6–12] . Prevention of human T . cruzi infection has traditionally relied on residual insecticide spraying campaigns and routine screening of blood-bank donors [2] . The two drugs ( nifurtimox and benznidazole ) registered for treatment of human infection with T . cruzi since the late 1960s and early 1970s were shown to be especially effective in young age groups during the acute and early chronic phase regardless of transmission mode [3 , 13–18] . Unfortunately , both nifurtimox and benznidazole cause adverse drug-related reactions ( ADR ) of various types , frequency and severity which increase with increasing patient’s age and reduce treatment compliance and effectiveness [19–22] . Benznidazole frequently causes mild or moderate dermatitis that respond well to antihistamines; low-dose oral glucocorticoids are less frequently needed . The rare cases presenting severe exfoliating dermatitis , dermatitis combined with fever and lymphadenopathy , and bone marrow suppression prompt immediate treatment discontinuation and intensive medical care [16 , 19 , 22 , 23] . Therefore , chemotherapeutic programs of human T . cruzi infection ideally should provide access to diagnosis and treatment as early as possible during the life course , minimize the occurrence of ADRs leading to reduced medication adherence , and avert eventual life threats posed by severe ADRs in the absence of timely medical attention . Less than 1% of patients infected with T . cruzi have access to parasiticidal treatment [24] . Most populations living under poor and marginalized conditions often lack access to diagnosis and treatment of T . cruzi infection ( and other neglected diseases ) and are unaware of their condition [25] . They also ignore disease consequences and the opportunities and limitations of current therapies . Barriers to treatment are multiple and include lack of training on ADR management , misconceptions on medication-related risks , reluctance to provide treatment , wide fluctuations in medication availability , overburdened or distant healthcare services , socio-cultural aspects , and lack of effective vector control and surveillance [22 , 26–29] . The pioneering treatment programs of T . cruzi infection implemented by Médecines Sans Frontières ( MSF ) in various countries since 1999 demonstrated that the challenge was tractable with adequate resources and stringent procedures [21] . They proposed that “Etiological treatment of Chagas disease can and should be integrated at the primary health care level…” [21] . This recommendation has also been endorsed by others [3] and may be traced back to the Declaration of Alma Ata in 1978 [30] . The proposition is also related to the concept of innovative and intensified disease management ( IDM ) for NTDs that can be managed within the primary healthcare system through more intensive use of existing tools , as is the case of Chagas disease [31] . However , the challenge of how to address diagnosis and treatment of T . cruzi infection in resource-poor , remote rural settings through primary health care has yet to be developed and program effectiveness documented to meet the challenge of treating the sizable population of infected rural residents and correct health inequities . The primary healthcare model focuses on community participation and social empowerment [30] . Broad social participation of multiple sectors may augment the feasibility and sustainability of control interventions , more so in disperse rural areas including various cultural groups [32–34] . Community participation is expected to increase the coverage , effectiveness and sustainability of vector and disease control actions of Chagas and malaria [35–39] . For example , in a remote rural area of Santiago del Estero ( Argentine Chaco ) under sustained community-based vector control in the mid-1990s , 17 ( 65% ) of 26 T . cruzi-seropositive children aged up to 15 years of age treated with benznidazole or nifurtimox seroconverted to a negative status between 2 and 13 years posttreatment [39] . As part of a long-term program on the eco-epidemiology and control of Chagas disease in the Argentine Chaco , we developed , implemented and tested a strategy to increase access to diagnosis and treatment of human T . cruzi infection in sparsely populated rural sections of Pampa del Indio municipality including 13 villages . This strategy , based on strengthened primary healthcare attention and broad social participation , followed an initial phase of intensified vector control and surveillance across the municipality [40–42] . The underlying premise was that participatory methods and multisector cooperation combined with adequate external support would increase diagnosis-and-treatment coverage and adherence relative to historical local levels , manage ADRs effectively and achieve positive therapeutic responses , as we document in this paper .
The study protocol was approved and supervised by “Dr . Carlos Barclay Independent Ethical Committee for Clinical Research” , Buenos Aires , Argentina ( Protocol N° TW-01-004 ) . All clinical investigations were conducted according to the principles expressed in the Declaration of Helsinki . The Ministry of Health of Chaco province and local hospital authorities granted permission to conduct the activities herein described . All individuals participating in serosurveys and treatment accepted to do so and their parents or guardians provided written informed consent . When community meetings included indigenous residents , explanations were translated by an indigenous healthcare agent or by an appointed indigenous community member , and consent was obtained collectively and individually . The intervention was conducted in Pampa del Indio ( 25°55’S 56°58’W ) , Chaco , Argentina . The study area included 353 houses and a few public buildings grouped in 13 rural villages distributed over a 450 km2 section as described elsewhere [40] . The study area was inhabited by 1 , 187 people in 2007 ( 1 , 318 people as of 2012 , including 565 up to 18 years of age ) , most of which lived on a subsistence economy . The only existing medical facility was a first-level public hospital with four physicians; the primary healthcare system had 5 posts distributed across the study area , and there were 8 primary schools . The closest and farthest villages were at 6 and 37 km from the hospital through dirt roads , and there was no public transportation . Initial selection of the study area followed the recommendations of the Chagas disease control program of Chaco: lack of insecticide spraying campaigns over the previous 12 years; reportedly high infestation levels with T . infestans , and requests of vector control interventions made by local district and healthcare authorities . Following an initial survey which provided evidence of large house infestation and active parasite transmission , all inhabited house compounds were sprayed with pyrethroid insecticides in November-December 2007 [40 , 43] . House infestation with T . infestans was monitored every 4–6 months from 2007 to 2010 ( which revealed moderate pyrethroid resistance levels ) and annually thereafter; all houses found to be reinfested were selectively re-sprayed with insecticides after each survey [40] . House infestation at the time of human diagnosis and treatment ( September 2010-March 2011 ) was <1% and mainly occurred in peridomestic structures; none of the bugs collected were infected with T . cruzi as determined by microscopic analysis of feces at 400× . These conditions and an established vector surveillance system were taken as prerequisites to launch diagnostic and treatment activities . The diagnosis-and-treatment program included five successive phases: preparatory , participatory planning , capacity strengthening of local health personnel , diagnostic surveys , and treatment and follow-up . Treatment-related primary outcomes included treatment coverage and quality and therapeutic response . Treatment coverage was estimated as the percentage of seropositive patients up to 18 years of age at serodiagnosis that were treated with benznidazole relative to the number of seropositive patients in this age group who were eligible for treatment . Assessment of treatment quality was based on individual completion , medication adherence and ADR management . Adherence was evaluated at the periodic appointments through the percentage of benznidazole pills taken ( i . e . , number provided minus residual pills ) relative to those provided for each specific time period , and then averaged over the treatment period . Patients with three or fewer pill counts and those who were withdrawn from or abandoned treatment were excluded from adherence estimates . Evaluation of ADR management included the percentage of patients who presented ≥1 ADR and were able to complete treatment; secondary outcomes included duration of the severest episode and proper notification of the event . Therapeutic response ( i . e . , treatment failure ) was primarily evaluated through detection of T . cruzi DNA by kPCR and qPCR [3 , 13 , 47] in the subset of patients who completed the full treatment course and had ≥80% of medication adherence . Although conventional serodiagnosis was also performed , at least in the Gran Chaco region clearance of conventional anti-T . cruzi antibodies usually took several years even in young patients in the early chronic phase [12 , 16 , 17 , 21]; therefore , we did not expect conventional serology would provide early evidence of seroconversion at 180 dpt . For additional comparisons , historical levels of treatment coverage with benznidazole or nifurtimox at the local hospital were estimated through a retrospective search of clinical histories of residents from the study area during the previous five years , and through householders’ reports . We used Friedman’s two-way non-parametric analysis of variance to test for significant differences among repeated measurements of uremia , creatinine , AST , ALT and ALP performed before treatment and at 20 and 60 dpt , and Kendall’s K as an index of concordance . Fisher’s exact test or χ2 tests were used for investigating two by two contingency tables of independent data . Exact McNemar significance probabilities were calculated for paired data with small cell frequencies . The nominal level of statistical significance was set at a P value of 0 . 05 . All tests were performed using Stata 12 [50] .
The flow chart of the study population from census to follow-up is shown in Fig 1 . The overall coverage of serodiagnosis was 70 . 3% ( 395 of 562 ) in children and adolescents aged up to 18 years old , with a peak in the age group 5–9 years old ( Fig 2 ) . The seroprevalence of T . cruzi infection in residents up to 18 years old was 24 . 1% ( 95 of 395 ) . According to householders’ reports , the local health system had serologically examined for T . cruzi infection 16 ( 2 . 8% ) local residents aged up to 18 years before the current intervention ( Fig 2 ) . Ninety-five seropositive individuals were screened for eligibility ( Fig 1 ) . Reasons for lack of participation included refusal to initiate treatment , household emigration from the study area , impaired health conditions , and a previous treatment with benznidazole or nifurtimox . Treatment with benznidazole was initiated by all 78 eligible participants ( Fig 1 ) . Treatment coverage decreased with increasing age ( Fig 2 ) . Crude treatment coverage among all identified seropositive residents was 82% ( 78 of 95 ) . The mean ( ±SD ) age at treatment was 11 . 0±4 . 0 years old ( range , 3–19 ) , and 36 ( 46% ) treated patients were females . All seropositive children who initiated treatment were asymptomatic and displayed the indeterminate form of Chagas disease at baseline except one ( of 71 patients examined by electrocardiography ) with a congenital arrhythmia unrelated to Chagas . Hematocrit , hemoglobin , platelets , white cell counts and liver enzymes usually were within normal limits , except one case of anemia and 27 with eosinophilia ( ≥10% ) ; these laboratory results did not prevent the initiation of treatment . Friedman’s test showed statistically significant changes over 0 , 20 and 60 dpt in creatinine levels ( P = 0 . 026 , K = 0 . 502 ) , AST ( P = 0 . 012 , K = 0 . 485 ) , ALP ( P < 0 . 001 , K = 0 . 641 ) , and marginally significant changes in uremia ( P = 0 . 077 , K = 0 . 420 ) and ALT ( P = 0 . 082 , K = 0 . 421 ) . Although most laboratory tests remained within normal limits during treatment , three patients presented a ≥2× increase of ALP at 20 dpt coinciding with ADR episodes whereas other three showed elevated levels of ALT and ADRs . Eleven patients had a ≥2× increase of ALP at 60 dpt , but most of them had no ADR . No patient interrupted treatment due to laboratory abnormalities . Treatment was completed by 66 ( 85% ) of the 78 patients enrolled in the study ( Fig 1 ) . Completion rates decreased from 100% among children <5 years to 85% among young people aged 15–19 years ( Fig 2 ) . Individuals who did not complete treatment tended to be older ( 12 . 4±6 . 0 years , range 9–17 ) and had a balanced gender distribution ( 50% ) . Medication adherence across patients who completed the full treatment course averaged 97% ( range , 80–100% ) ( Table 1 ) . Among the 12 patients who did not complete treatment , four took medication for 25–28 days ( range of adherence , 70–100% ) ; other four took it for 4–17 days , and no data were provided by four patients . For comparison , the local health system had treated with benznidazole only two T . cruzi-seropositive children residing in the study area over the previous five years . A total of 24 ( 32%; 95% CI: 21–43% ) patients presented ≥1ADR ( Table 1 ) , and 20 ( 83% ) reported it through the agreed mechanisms . Table 1 only includes the severest presentation for the six patients with 2 or 3 ADR episodes . The mean age of patients showing ≥1 ADR ( 12 . 0 years , 95% CI: 10 . 3–13 . 7 ) was not significantly different from that of patients showing no ADR ( 10 . 5 years , 95% CI: 9 . 4–11 . 6 ) . On average , the first ADR appeared at 13 . 3 dpt initiation ( 95% CI: 10 . 8–15 . 8; range , 4–29 dpt ) . Thirty-eight ADR episodes were recorded , including 22 mild , 9 moderate and 4 severe presentations , and 3 dermatological ADRs whose severity could not be established because of late reporting ( Table 1 ) . Twenty-one ( 88% ) patients with ≥1 ADR displayed maculopapular exanthema , alone or combined with mild headaches or an increasingly severe arthralgia/myalgia . No Lyell or Stevens-Johnson syndromes and polyneuritis were recorded . Mild and moderate ADRs were managed in an out-patient basis; physicians evaluated the patients and eventually administered antihistamines , paracetamol and ibuprofen . Temporary dose reductions were indicated to four patients . Benznidazole was temporarily suspended for an average of 4 days in six patients showing mild or moderate exanthema and mild to severe arthralgia/myalgia or moderate headaches . Treatment withdrawal ( followed by short-term hospitalization and administration of corticoesteroids and antihistamines ) was prescribed to four patients displaying repeat ADR episodes that did not respond to symptomatic medication ( age range , 6–17 years ) : three had a severe exanthema ( one with fever ) and one a moderate , prolonged exanthema and headaches . Duration of the severest ADR episode took on average 3 . 4 days ( range , 1–7 days ) . Medication adherence varied very little among ADR types and levels . Eighteen ( 75% ) patients with ≥1ADR were able to complete treatment . Treatment completion rates significantly decreased with increasing ADR severity from 90–92% among patients with either no ADR or mild exanthema ( including headaches and dizziness ) to 0% among the three patients with a severe exanthema ( Fisher’s exact test , P = 0 . 011 ) , after pooling no or mild ADRs versus moderate or severe ADRs to avoid small cell frequencies ( Table 1 ) . One patient who failed to report an ADR ( exanthema ) abandoned treatment without medical indication and prompted other four family members under treatment to do so despite they had no ADR . Among patients having another household member under treatment , the relative odds of presenting at least one ADR was significantly and positively associated with having another household member with an ADR ( OR = 2 . 57; 95% CI: 1 . 02–7 . 28; exact McNemar significance , P = 0 . 043 ) . The therapeutic response to treatment as determined by kPCR and qPCR is shown in Table 2 . Before treatment , 40 ( 65% ) patients were co-positive by both PCRs , 15 ( 24% ) were co-negative , and discordant results occurred in 7 ( 11% ) qPCR-positive and kPCR-negative patients; the performance of both PCRs differed significantly ( exact McNemar significance , P = 0 . 016 ) . During treatment , only one ( 2% ) patient was kPCR-positive and qPCR-negative whereas the remainder was co-negative . Immediately after treatment at 60 dpt , 56 ( 97% ) patients were co-negative , 1 ( 2% ) was kPCR-negative and qPCR-positive , and 1 ( 2% ) showed the reverse discordant pattern . This patient`s parasite burden decreased from 20 Pe/mL before treatment to 0 . 44 at 60 dpt and was negative thereafter . At 180 dpt , the only patient who was co-positive had shown T . cruzi DNA amplification at 20 dpt by kPCR ( i . e . , treatment failure ) and refused the offer for new treatment with nifurtimox . The relative frequency of kPCR- or qPCR-positive results before treatment highly significantly declined by 180 dpt ( exact McNemar significance , P < 0 . 0001 ) . Median parasite load among kPCR- or qPCR-positive patients before treatment drastically fell from 1 . 4 Pe/mL to undetectable levels at 20 dpt and thereafter remained much lower than before treatment among the three patients positive by either PCR ( Table 2 ) . Of the 46 patients who were qPCR-positive before treatment , only two were positive at 60 ( subsequently negative ) or 180 dpt; the remainder was two ( 42 ) or three ( 38 ) times negative over 20–180 dpt . Of 16 patients initially qPCR-negative , none was positive over 20–180 dpt . The treated patients who participated in the serological follow-up at 60 or 180 dpt showed virtually no decay in paired optical densities by the Chagatest ELISA between 0 and 60 dpt ( mean percent absolute reduction , 5 . 5%; 95% confidence interval , CI , 1 . 3–9 . 8 ) or between 0 and 180 dpt ( mean , 0 . 5%; CI , -1 . 6–1 . 7% ) . Only one patient had a drop in optical density levels below the Chagatest cutoff value . Using recombinant ELISA , mean percent absolute reductions between 0 and 60 dpt ( 6 . 6%; CI , 2 . 9–10 . 4% ) and between 0 and 180 dpt ( mean , 7 . 1%; CI , 2 . 3–11 . 8% ) were slight at most , with only one patient having a >60% drop in optical densities .
Our study demonstrates that diagnosis and treatment of T . cruzi infection in remote , impoverished rural areas can be effectively addressed through strengthened primary healthcare attention and broad social participation combined with adequate external support . This strategy took advantage of locally available resources; secured high levels of diagnostic and treatment coverage and medication adherence , effectively managed ADRs at village level under carefully administered protocols , and provided early evidence of positive therapeutic responses except in one case . Although the efficacy of benznidazole has been firmly established in hospitals or specialty clinics [15 , 16] , especially among young patients , the issues of treatment access , adherence and effectiveness in remote , hyperendemic rural settings have received little attention . Broad social participation implied a multi-stakeholder agreement that included the affected communities and other sectors . The community workshops were essential to raise awareness of Chagas disease characteristics and consequences , diagnostic and treatment opportunities , and provided an appropriate context to reach an agreement on intervention details adapted to local circumstances . A key output was the two-sided commitment for implementing interventions and building of mutual trust . Compliance with dates for the expected return of serological results ( unlike in other health interventions reported by householders ) facilitated further community involvement with treatment and follow-up activities . The local health system faced critical constraints ( e . g . , medical personnel and vehicle ) at the time of the interventions . These facts , combined with considerable distances between rural villages and the local hospital , led to a centralized health service delivery model to which the rural population had poor access . The initial workshop outputs identified that most activities had to be conducted at the residents’ villages . Strengthening the capacity of rural healthcare agents and medical personnel was essential: the former undertook monitoring of medication adherence and ADRs , whereas the in situ participation of physicians was restricted to treatment indication , clinical exams and improved ADR management . In remote areas of Africa , transference of treatment initiation and monitoring of HIV patients to local nurses improved the access to and quality of health care and cost-effectiveness of integrated interventions [51 , 52] . Benznidazole was available at the time of our program and timely provision was carefully planned , but these usually are major issues elsewhere . As a direct repercussion of the current program , spontaneous treatment demand at the local hospital increased substantially over subsequent years and included adult patients . The degree of diagnostic coverage attained among rural residents up to 18 years old was substantial ( 70 . 3% ) and paralleled or exceeded levels recorded in rural communities of the Argentine Chaco [6 , 39] and by the local health system . The enhanced diagnostic coverage was partly related to the prospects of receiving treatment and eventual welfare benefits derived from being seropositive for T . cruzi . Treatment completion rates were likely enhanced through the contribution of dedicated staff supervising the onset of benznidazole administration and ADR follow-up; no incentives for treatment initiation , completion and follow-up were given . Medication adherence was very high and similar to the levels recorded in urban or rural populations through other approaches [19 , 21 , 22] . The healthcare agent- and parent-based monitoring system was key to early identify patients showing ADRs and breaches in adherence . Likewise other benznidazole trials in children [19 , 21] , most ADRs occurred during the first two weeks after onset of treatment; close patient monitoring during this period is therefore indicated . No significant age-related increase of the occurrence of an ADR was detected , unlike in studies covering a wider age range [20 , 26] . The significant household aggregation of ADRs suggests putative genetic or environmental factors acting on a familial level that may possibly modify the bioavailability of or response to benznidazole . Elevated levels of ALT or ALP during or immediately after benznidazole treatment occurred in several patients , but they were not as important as to interrupt treatment ( rarely exceeded baseline levels by 3× ) nor were laboratory abnormalities associated with the occurrence of an ADR . The most frequent benznidazole-related ADRs were dermatological ( usually of mild or moderate severity ) as in several other studies and locations , although their frequency tends be quite variable [12 , 19 , 21 , 22] . Most mild or moderate ADRs were successfully managed with symptomatic medication , dose reduction and temporary suspension of benznidazole intake , whereas the four cases with severe ADRs required treatment withdrawal , short-term hospitalization and symptomatic medication . Treatment completion was significantly and inversely related to ADR severity , again supporting the need of close monitoring and adequate ADR management as a means of increasing completion rates and preventing treatment abandonment in other family members . qPCR was significantly more sensitive than kPCR to detect T . cruzi DNA before treatment and evidenced the rapid fall of circulating parasites as early as three weeks after onset of treatment [20 , 48] . Evidence from several trials supports that qPCR is an early marker of treatment failure [13 , 31 , 46–48] . The proportion of T . cruzi-seropositive patients initially qPCR-positive ( 65% ) was very close to the range recorded in several other studies targeting the same age group and early chronic infections [18 , 53 , 54] . Moreover , all children [22] and the great majority of adult patients [13] treated with benznidazole for 60 days remained qPCR-negative over 1–1 . 5 years of follow-up after effective treatment , hence suggesting that under such conditions recurrence of parasitemia appears to be rare but other patterns have been reported [e . g . , 18] . In our study , the only patient with persistently positive kPCR or qPCR tests ( before and at 20 and 180 dpt ) was a 12-year-old child with rather limited adherence ( 80% ) who had no travel history outside the study area and resided in a non-infested house ( i . e . , a treatment failure ) . However , a sizable fraction of adult individuals under a partial treatment course with benznidazole have shown positive therapeutic responses [55] . A second patient with very much reduced parasite burden from 0 to 60 dpt ( but still qPCR-positive ) was subsequently qPCR-negative , as recorded by others [46] . Most important , all patients but two were subsequently qPCR-negative on two or three occasions over 20–180 dpt . Although we cannot exclude whether longer follow-up times would allow some of the molecular tests to yield a positive result , the individual time patterns of qPCR including two or three negative results in a row are more compatible with a process of parasite clearance in young patients in the early chronic phase of infection [cf 14 . 17] . Intensified vector control and systematic surveillance was deemed a prerequisite for launching the diagnosis-and-treatment program because fast house reinfestation after insecticide spraying frequently causes new human infections in rural areas of the Gran Chaco [5 , 8 , 10] and elsewhere [18 , 56] . In the face of moderate pyrethroid resistance , suppressing house infestations demanded recurrent insecticide applications [40] and delayed treatment activities . However , this prolonged process laid the foundations for subsequent participatory activities; averted the chance of new vector-mediated human infections , and determined that child infections most likely had been acquired at least 3–4 years before treatment ( i . e . , early chronic infections ) . Our study had several limitations . Generalizability of the current strategy to other areas depends on the existence of a primary healthcare service . Medication adherence was measured through pill counts in the absence of a practical assay for establishing serum benznidazole levels at the study setting . In the absence of a well-established biomarker of early cure [26 , 47] , the time-limited follow-up of patients up to 180 dpt precluded us from using seroconversion to a negative status as a primary endpoint . Seroconversion usually takes several years even in young residents from the Gran Chaco in the early chronic phase [12 , 16 , 17 , 21] , depending on the time span between primary infection and treatment and other ill-defined determinants , which explains the nearly stable pattern of ELISA optical densities before and after treatment . We note , however , that results elsewhere in central Brazil [14 , 15] , Guatemala and Honduras [21] and Colombia [18] showed quite diverse patterns of conventional antibody clearance rates in response to benznidazole treatment in young patients <15 years old . Losses to follow-up via PCRs at 180 dpt included 17 ( 26% ) of 66 patients who completed treatment and may bias estimates of treatment effect size . Cohort attrition over the follow-up is typical in trials conducted in rural locations , more so in remote areas with migrant populations . Whether T . cruzi-infected children treated with benznidazole will show improved long-term clinical outcomes is a crucial question that merits further research . Conclusions from this study may not be extrapolated to adult patients who have more frequent and severe ADRs with a different time pattern [22] . Our study links community participation to a health outcome improvement [57 , 58] . Community participation in remote rural settings is essential for treatment programs [32] . Increasing access to high-quality serodiagnosis and treatment of marginalized rural populations , combined with effective vector control and surveillance in the affected regions , is ethically imperative . | Less than 1% of patients infected with Trypanosoma cruzi have access to parasiticidal treatment with the two available drugs , including millions of patients in the early chronic phase who would greatly benefit from treatment . Furthermore , rural populations living under poor and marginalized conditions usually have little or no access to chemotherapeutic programs of Chagas and other neglected tropical diseases . Barriers to treatment range from misconceptions on medication-related risks to socio-cultural aspects , among others . This study addressed the challenge of diagnosis and treatment of T . cruzi infection in resource-poor , remote rural settings of the Argentine Chaco where 24% of residents up to 18 years of age were infected . The underlying premise was that participatory methods and multisector cooperation would increase program effectiveness . Our results show that diagnosis and treatment of T . cruzi infection there can be effectively addressed through strengthened primary healthcare attention and broad social participation with adequate external support , following an initial phase of intensified vector control and surveillance across the municipality . This strategy secured high diagnosis-and-treatment coverage and adherence; effectively managed adverse drug-related reactions; provided early evidence of a positive therapeutic response , and may stimulate healthcare services in the affected regions to aggressively bridge the treatment gap . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"health",
"care",
"health",
"care",
"providers",
"protozoans",
"pharmaceutics",
"neglected",
"tropical",
"d... | 2017 | Improving access to Chagas disease diagnosis and etiologic treatment in remote rural communities of the Argentine Chaco through strengthened primary health care and broad social participation |
Interleukin ( IL ) -22 , an immune cell-derived cytokine whose receptor expression is restricted to non-immune cells ( e . g . epithelial cells ) , can be anti-inflammatory and pro-inflammatory . Mice infected with the tapeworm Hymenolepis diminuta are protected from dinitrobenzene sulphonic acid ( DNBS ) -induced colitis . Here we assessed expulsion of H . diminuta , the concomitant immune response and the outcome of DNBS-induced colitis in wild-type ( WT ) and IL-22 deficient mice ( IL-22-/- ) ± infection . Interleukin-22-/- mice had a mildly impaired ability to expel the worm and this correlated with reduced or delayed induction of TH2 immunity as measured by splenic and mesenteric lymph node production of IL-4 , IL-5 and IL-13 and intestinal Muc-2 mRNA and goblet cell hyperplasia; in contrast , IL-25 increased in the small intestine of IL-22-/- mice 8 and 12 days post-infection compared to WT mice . In vitro experiments revealed that H . diminuta directly evoked epithelial production of IL-25 that was inhibited by recombinant IL-22 . Also , IL-10 and markers of regulatory T cells were increased in IL-22-/- mice that displayed less DNBS ( 3 mg , ir . 72h ) -induced colitis . Wild-type mice infected with H . diminuta were protected from colitis , as were infected IL-22-/- mice and the latter to a degree that they were almost indistinguishable from control , non-DNBS treated mice . Finally , treatment with anti-IL-25 antibodies exaggerated DNBS-induced colitis in IL-22-/- mice and blocked the anti-colitic effect of infection with H . diminuta . Thus , IL-22 is identified as an endogenous brake on helminth-elicited TH2 immunity , reducing the efficacy of expulsion of H . diminuta and limiting the effectiveness of the anti-colitic events mobilized following infection with H . diminuta in a non-permissive host .
Interleukin ( IL ) -22 , a member of the IL-10 family , is produced predominantly by innate ( NK cells ( NK22 ) , γδ T cells , innate lymphoid cells type 3 ( ILC3s ) and adaptive ( CD4+ Th22 and Th17 , CD8+ T cells ) immune cells: a non-immune source of IL-22 has not been described . The heterodimeric IL-22 receptor consists of the IL-10R2 subunit and the unique IL-22R1 subunit , and is restricted to non-hematopoetic cells ( e . g . hepatocytes and epithelium of the gastrointestinal tract ) [1] . Thus , IL-22 is an immune cell-derived mediator that acts exclusively on non-immune cells and as such is an attractive target for therapeutic intervention [2 , 3] . Data from studies of the gastrointestinal tract suggest that the role of IL-22 is contextual , with beneficial or detrimental affects depending on the nature of the disease or immune activity being assessed . For example , IL-22-/- mice are more susceptible to colitis induced by dextran sodium sulfate ( DSS ) [4] and retinoic acid suppression of DSS-induced colitis was associated with increased IL-22 [5] . Similarly , local delivery of the IL-22 gene attenuated the spontaneous colitis that develops in T cell receptor ( TCR ) -α knockout ( KO ) mice [6] and that evoked by transfer of naïve CD45RBhi T cells into RAG2-/- mice [4] . However , IL-22 , mobilized by IL-23 , was implicated in the exaggeration of murine colitis induced by anti-CD40 activation in RAG1-/- mice [7] . Fewer IL-22+ cells have been described in inflamed tissue from patients with ulcerative colitis compared to healthy individuals [8] . In contrast , ILC3 from patients with mild-moderate ulcerative colitis were reported to have increased IL-22 production [9] . This duality of IL-22 function extends beyond the gut . Interleukin-22 can promote hepatocyte survival in acute mouse models of liver damage [10] , while IL-22 recruitment of Th17 cells has been implicated in chronic liver inflammation of hepatitis B-infected individuals [11] . Pro- and anti-inflammatory roles have been described for IL-22 in murine models of arthritis [12]; for example , IL-22 was implicated in the enhancement or suppression of collagen-induced arthritis in mice co-treated with the parasitic nematode-derived molecule , ES-62 [13] . The role of IL-22 following infection is equally diverse , where it has been shown to protect mice from infection with Citrobacter rodentium and Salmonella enterica [14] , but appears not to affect the outcome of infection with Mycobacterium avium [15]; susceptibility to Salmonella has been reported [16] . The route of pathogen entry into the body can be important , IL-22 acting downstream of IL-23 , promoted resistance against intragastrically or intravenously delivered Candida albicans [17] , but played no role in the response to cutaneous C . albicans [18] . Two independent studies demonstrated roles for IL-22 in the intestinal pathophysiology associated with infection with Toxoplasma gondii [15 , 19] . With respect to infection with helminth parasites , Wilson et . al . found no role for IL-22 in the murine response to Schistosoma mansoni [15] , whereas goblet cell hyperplasia and mucin secretion , a key effector in the gut , was driven by IL-22 following infection with nematodes [20] . Increased IL-22 has been demonstrated in individuals with established hookworm infection although its function was not defined [21] . A report of self-infection with the nematode parasite Trichuris trichiura to treat ulcerative colitis documented increased numbers of CD4+IL22+ cells [22] . Infection with the rat tapeworm , Hymenolepis diminuta , protects mice from colitis induced by intra-rectal ( i . r . ) instillation of the haptenizing agent , 2 , 4-dinitrobenzene sulphonic acid ( DNBS ) [23] . Given the pivotal role that IL-22 can play in immune-stromal cell communication and the disparite data on this cytokine in the response to infection ( and general lack of data in relation to helminths ) and regulation of inflammation , the current study assessed the impact of the absence of IL-22 in ( 1 ) the expulsion of H . diminuta from its non-permissive mouse host and the concomitant immune response , and ( 2 ) whether the anti-colitic effect of infection with H . diminuta was modified .
The role of IL-22 in modifying the host response to infection with helminth parasites appears to be determined by the nature of the infection . For example , worm burden and granuloma size is not different in schistosoma-infected WT and IL-22-/- mice [15] , whereas IL-22 was important in the goblet cell hyperplasia and mucin secretion response following infection with the intestinal nematodes , Trichuris muris and Nippostrongylus brasiliensis [20] . The tapeworm H . diminuta is unique amongst helminths that infect the intestine as it does negligible , if any , damage to the host: it lacks a tissue migratory phase and the absence of hooks on the scolex means it is not abrasive . IL-22-/- mice displayed a slight delay in the kinetics of expulsion of H . diminuta: only 22% ( 2/9 mice ) of infected IL-22-/- mice had expelled H . diminuta by 8 days post-infection ( dpi ) compared to 55% ( 5/9 mice ) of WT mice ( Fig 1 ) ; at this time-point 33% of infected IL-22-/- mice harboured 3 or 4 worms , burdens not observed in WT mice . At 12 dpi , H . diminuta had been completed expelled from WT and IL-22-/- mice , suggesting that while IL-22 signaling promotes a rapid anti-H . diminuta response the duration of infection is not prolonged in the absence of this cytokine . Mobilization of TH2-type cytokines ( i . e . IL-4 , IL-5 and IL-13 ) is a hallmark of the immune response following infection with parasitic helminths [24] . Consistent with previous findings [25] , mitogen stimulation of splenocytes or mesenteric lymph node ( MLN ) cells from WT mice resulted in increased IL-4 , IL-5 and IL-13 by 4-dpi ( Fig 2A and 2B ) , declining to control levels by 12-dpi . Time-matched analyses revealed reduced levels of the 3 cytokines from MLN and spleen of IL-22-/- mice on day 4-dpi compared to WT mice , that rebounded to match or exceed those of WT mice by day 8-pdi ( the exception being IL-13 production by MLN cells ) ( Fig 2A and 2B ) . By 12-dpi there were no differences in splenic and MLN-derived IL-4 , IL-5 or IL-13 in infected WT and IL-22-/- mice . Measurement of the TH1 cytokine IFN-γ from conA-stimulated splenocytes revealed no differences between WT and IL-22-/- mice over the 12-day infection period ( S1 Fig ) . In addition , qPCR revealed reduced expression of IL-4 , IL-10 and IL-25 mRNA in intestinal tissue from infected IL-22-/- mice compared to WT animals at 4-dpi , with a rebound heightened expression in all 3 cyokines by 8-dpi , which unlike the spleen and MLN was extended until 12-dpi ( Fig 2C ) ( end of experiment ) . This delay in the production of key TH2 effector cytokines parallels the delay in expulsion of H . diminuta from IL-22-/- mice and the events are likely to be causally linked . These data align with the requirement for IL-25 in the expulsion of nematode parasites from mice [26–28] . Fascinatingly , and in accordance with IL-22’s dual functions [13] , the diminished TH2 responses in IL-22-/- H . diminuta-infected mice suggests an important role for innate immunity early in the response to helminths and additional studies are needed to precisely define this . Interleukin-4 has been implicated in the regulation of goblet cell hyperplasia following infection with helminth parasites [29] . Indeed , mucin synthesis and release are important , often critical , effector responses against enteric helminths [30] and goblet cell hyperplasia follows the kinetics of H . diminuta expulsion from WT mice [31] . Four dpi mRNA for the secreted mucin , Muc-2 , was increased in the small intestine of infected WT and to a lesser extent in IL-22-/- mice: and while Muc-2 mRNA expression declined in the intestine of WT mice , in IL-22-/- mice the elevated Muc-2 expression was maintained at 8-dpi , paralleling the kinetics of H . diminuta expulsion ( Fig 3A ) . The Muc-1 gene encodes a transmembrane bound mucin; little is known of its function [32] . Muc-1 mRNA was significantly upregulated in H . diminuta-infected IL-22-/- mice at 8- and 12-dpi and it is tempting to speculate that this might compensate for the reduced Muc-2 signal at 4-dpi in these mice ( Fig 3A ) . Rats , the natural definitive host for H . diminuta , infected with 5 cysticercoids show no increase in Muc-2 mRNA , whereas a 50 cysticercoid oral inoculum resulted in increased Muc-2 mRNA , and ≤15 worms established in the gut [33] . Histochemical staining revealed increased numbers of mucus-containing goblet cells in the small intestine of H . diminuta-infected WT mice ( Fig 3B ) [31]; however , intestine from infected IL-22-/- mice displayed no significant increase in goblet cells at 4-dpi ( Fig 3B ) . The reduced Muc-2 expression and parallel changes in goblet cell numbers in IL-22-/- mice could contribute to the increased worm burden observed at 8-dpi , while maintenance of the Muc-2 signal and sustained goblet cell numbers ( Fig 3C ) may allow for these mice to catch-up with WT animals , fully expelling H . diminuta by 12-dpi . However , neither Muc2 mRNA nor goblet cell numbers are substantially increased beyond WT levels at 12-dpi , despite increased IL-4 and IL-25 mRNA in the small intestine , suggesting that once the parasite has been eradicated ( see Fig 1 ) , regulatory mechanisms come into play to dampen a mucus/goblet cell response . Interleukin-22 has been implicated in the barrier function of the gut , especially the secretion of anti-microbial factors and mucin [34] , and while this can be a direct effect , the diminutation of IL-4 or IL-13 production in the IL-22-/- mice could contribute to the perturbation of mucin and goblet cell regulation following infection with helminth parasites . Intestinal mast cell hyperplasia can accompany infection with nematodes [24] , but c-Kit immunostaining revealed comparable numbers ( and distribution ) of mast cells in WT and IL-22-/- mice ( S2 Fig ) . These data suggest a limited , if any , role for mast cells in the current study but an in-depth analysis is required before definitive statements on the role of mast cells ( with or without IL-22 ) in the response to H . diminuta can be made . Juxtaposing the facts that the epithelium is a target for IL-22 [1] and epithelium-derived factors are important in shaping the immune response and the outcome of infection [35] , the impact of the absence of IL-22 on the mobilization of regulatory immune factors/cells was assessed following infection with H . diminuta . The observation of increased IL-25 mRNA in the jejunum of IL-22-/- mice at 8- and 12-dpi with H . diminuta ( Fig 2C ) suggested that IL-22 serves as a brake on the synthesis of tissue ( i . e . epithelial ) -derived cytokines elicited in response to infection with helminth parasites . The increase in IL-25 mRNA in the IL-22-/- mice could be due to increased presence of the parasite and not the IL-22-/- deficiency per se . To test this , WT mice were infected with 5 or 10 H . diminuta , and while the latter did lead to increased spleen cell number and TH2 cytokine output , there were no differences in worm burden or intestinal IL-25 mRNA levels between the two infection paradigms at 8-dpi ( S3 Fig ) . Thus , a higher antigenic load is not responsible for the increased IL-25 response but rather this is attributable to the absence of IL-22 . Focusing on IL-25 , murine IEC4 epithelial cells were exposed to a single H . diminuta ( scolex and ~2 cm of strobila ) ± recombinant IL-22 . Levels of IL-25 protein and mRNA expression were determined in supernatant and Trizol-treated cells , respectively . The epithelia spontaneously produced IL-25 that was significantly increased by H . diminuta , and in both cases IL-22 reduced IL-25 production ( Fig 4A ) , correlating with mRNA levels ( Fig 4B ) . To our knowledge this is the first time that IL-22 suppression of IL-25 production in the context of a parasitic helminth infection has been shown , underscoring the role of IL-22 in moulding the host response following infection . In addition , using the reductionist approach of culturing a single H . diminuta scolex with epithelial cell lines , we found epithelia from the non-permissive mouse host produced IL-25 , IL-33 and TSLP ( mRNA and protein ) and that the rat ( permissive host ) IEC . 6 cell line failed to show this alarmin response to the worm [36] . Of note , qPCR revealed a trend towards increased IL-33 and thymic stromal lymphopoietin ( TSLP ) mRNA expression in the jejunum of H . diminuta infected mice ( S4 Fig ) ; others have shown differential regulation of IL-25 , IL-33 and TSLP following infection with helminth parasites [35] . In addition to its role as a TH2-polarizing cytokine , IL-25 inhibition of trinitrobenzene sulphonic acid ( TNBS ) -induced colitis in mice may involve alternatively activated macrophages ( AAMs ) [37] , and markers of AAMs are increased in the gut of H . diminuta-infected mice [38] . Extrapolating from this , the increased IL-25 production from epithelia exposed to H . diminuta and the highly significant increase in IL-25 mRNA in the parasitised gut of IL-22-/- could result in increased mobilization of immunoregulatory cells and suppression of concomitant disease in the infected mice . IL-10 synthesis follows infection with H . diminuta infection and is an important anti-inflammatory cytokine in mice and humans [23] . The increased levels of IL-10 , Foxp3 and markers of AAMs ( i . e . arginase-1 and Fizz1 ) mRNA found in the small intestine of H . diminuta-infected Balb/c mice [25] , suggests expansion of innate and adaptive regulatory cells . These data were confirmed and extended here and , moreover , gut levels of IL-10 mRNA ( Fig 2C ) and stimulated IL-10 from splenocytes ( Fig 5A ) and MLN cells ( Fig 5B ) were significantly increased at 8- and 12-dpi in IL-22-/- mice compared to WT mice . Macrophages can be an important source of IL-10 in response to helminth and microbial antigens [39 , 40] . However , macrophages differentiated from the bone-marrow of IL-22-/- mice had a normal capacity to produce IL-10 in response to H . diminuta antigen or LPS ( S5 Fig ) ; thus , we speculate that the increased IL-10 observed in MLN and splenocytes at the later time-points of infection in IL-22-/- mice is from T cells , or potentially B cells [23 , 41] . We have shown a variable increase in Foxp3 mRNA in the small intestine of H . diminuta-infected Balb/c mice [25] . Despite the likelihood of IL-22-Foxp3 cross-regulation [42] little is known of the putative interaction of these two factors following infection with helminth parasites . Increased Foxp3 mRNA was observed in the small intestine of IL-22-/- H . diminuta-infected mice compared to WT animals ( Fig 5C ) , supporting the notion that IL-22 serves as a brake on immunoregulatory cell mobilization; however , immunoblotting with extracts of small intestine failed to show a consistent increase in Foxp3+ cells , which may reflect sensitivity of this assay as compared to qPCR ( S6 Fig ) . Moreover , while CD4+Foxp3+ splenocytes were increased following infection ( 8-dpi ) there were no differences between WT and IL-22-/- mice . ( Fig 5D and 5E ) . The reason for the discrepancy between small intestine and spleen is unclear but it underscores the complexity of immunoregulation and the need to precisely define events in both time and space as they relate to the host response to infection . In addition , expression of Foxp3 does not unequivocally identify a cell with immunosuppressive capacity [43] , and so IL-10 may be more important than Foxp3 in immunoregulation in this helminth-rodent model system [23] . Although the interaction of IL-25 and Foxp3 expression was not pursued , the association is noteworthy , given data showing lower numbers of Tregs in IL-25-/- mice [44] , increases in antigen-specific IL-22+ T cells concomitant with fewer Foxp3+ T cells in an individual with ulcerative colitis infected with T . trichura [22] , and that NOD mice treated with IL-25 have increased numbers of Tregs [45] . Thus , one can speculate that the increase in IL-25 in IL-22-/- mice could mediate the increase in Foxp3 and hence Tregs . The role of IL-22 in controlling the mobilization and activity of immunoregulatory cells is not well understood and in addition to considering Tregs , the putative impact of IL-22 on B cells should not be overlooked: for example , successful treatment of tuberculosis correlated with , but was not functionally linked , to increased IL-22 production and a reduced frequency of putative regulatory CD5+CD1d+ B cells [46] . Based on the changes observed in IL-22-/- mice following infection with H . diminuta , we examined the impact of lack of IL-22 in ( a ) the outcome of DNBS-induced colitis and ( b ) the ability of infection with H . diminuta to reduce the severity of DNBS-induced colitis . IL-22-/- mice consistently developed less severe DNBS-induced colitis compared to WT mice , in all indicies measured: weight loss , colon length and macroscopic appearance , MPO activity ( indicative largely of neutrophil infiltration ) and the cumulative disease activity score ( DAS ) ( Fig 6 ) . This effect was compounded following infection with H . diminuta: infected IL-22-/- mice treated with DNBS showed minimal signs of disease and were often indistinguishable from control , non-DNBS-treated mice ( Fig 6 ) . These findings are in accordance with the reduced mobilization of IFNγ and neutrophils observed in T . gondii-infected IL-22-/- compared to WT mice [47] ( ( DNBS-induced colitis is considered a TH1-dominated disease and hence the balance of TH1 and TH2 immunity is important in disease severity [48] ) . Corroborating these macroscopic measures of disease activity , histological analyses revealed that IL-22-/- mice had less DNBS-induced histopathology compared to WT mice , and only very minor damage was observed in the colon of H . diminuta-infected IL-22-/- mice ( Fig 7A and 7B ) . Mitogen stimulation of splenocytes from WT or IL-22-/- mice infected with H . diminuta revealed increased IL-10 production compared to uninfected mice , both naïve and DNBS-treated . Cells from infected DNBS-treated IL-22-/- mice produced , on average , more IL-10 than WT mice , but this did not reach statistical significance ( p = 0 . 2 ) ( Fig 7C ) . In contrast , splenic production of IL-17 , while increased by DNBS , was not significantly different between WT and IL-22-/- mice ± infection with H . diminuta ( Fig 7D ) . Juxtaposing these data with those from H . diminuta-infected naïve IL-22-/- ( Figs 2–5 ) , it is likely that the increase in IL-10 , IL-25 and putative regulatory T cells ( i . e . increased jejunal Foxp3 mRNA ) enhances the anti-colitic effect of infection with H . diminuta in mice lacking IL-22 . It has been reported that IL-22 protects female Balb/c mice from TNBS ( 3 mg , 5 days ) -induced colitis [49] ( yet others found no increase in IL-22 mRNA in TNBS-treated animals [50] ) . Given the structural similarity of DNBS and TNBS how can these disparate roles of IL-22 be reconciled ? Differences in the sex of mice , the duration of the disease and the natural microbiota of the mice could , at least in part , underlie the opposing findings of the two studies . Also , the protective effect of IL-22 in TNBS-colitis was based on administration of a neutralizing antibody and not genetic knockout of the IL-22 gene , raising the possibility of non-IL-22 effects of the antibody . Again , the point arises that the beneficial versus detrimental impact of manipulating IL-22 as a therapy will be contextual . IL-22-/- mice have increased susceptibility to dextran sodium sulfate ( DSS ) -induced colitis [4] and hence the findings in the DNBS model were somewhat surprising . We confirmed that the IL-22-/- mice used here had heightened responsiveness to DSS ( S7 Fig ) . The increased severity of DSS-induced colitis in IL-22-/- mice has been linked to a pro-colitiogenic microbiota [4] . To address this , a published protocol [51] was used to blend the microbiotas between WT and IL-22-/- mice prior to DNBS treatment . The severity of colitis in IL-22-/- mice with their natural microbiota and those who acquired microbiota from WT mice was not different , and both had significantly less disease than WT mice ( S8 Fig ) . In contrast , all of the WT mice who acquired microbiota from IL-22-/- mice presented with severe DNBS-induced colitis , with a marked increase in the size of the cecum: these mice were the sickest of all the experimental groups ( S8 Fig ) . Thus , IL-22-/- mice may harbour a microbial pathobiont that is not important to DNBS-induced colitis in these mice but exaggerates disease in WT mice , somewhat analysis to the transmissibility of susceptibility to DSS by the microbiota from IL-22-/- mice [4] . Assessing the possibility that IL-22-/- could be deficient in anti-microbial peptides , qPCR revealed that this was not the case . In line with findings reported in intestinal bacterial infection increases in mRNA for β-defensin 1 , 2 and 3 was similar in WT and knock-out mice following infection with H . diminuta ( S9 Fig ) . Interestingly , unlike infection with C . rodentium that increased RegIIIβ and RegIIIγ in a IL-22-dependent manner [14] , infection with H . diminuta evoked only a transient increase in RegIIIβ but not RegIIIγ mRNA ( S9 Fig ) . Thus , the contribution of IL-22 to DNBS-induced colitis is not likely due to different microbiota rather it is a consequence of altered immunoregulation in the absence of IL-22 . The fact that IL-22-/- mice experience less DNBS-induced and greater DSS-induced colitis highlights important differences in disease pathogenesis . Up-regulation of IL-22 mRNA has been found in DSS- but not in TNBS-induced colitis [50] . In the gut , T cells , γδ T cells and ILC3s are major sources of IL-22 [52] . More recently neutrophils have been cited as a source of IL-22 [53] . However , the extent to which each cell is activated in colitis and by which stimuli ( i . e . cytokines vs . pattern-associated microbial patterns ) is not fully understood . Consequently additional efforts are required to unravel the role of IL-22 in a variety of model systems and in the context of varying microbiotas if extrinsic manipulation of IL-22 levels is to be considered a treatment for enteric disease . The situation is complicated further by the recent demonstration that IL-25-/- mice are protected from DSS-induced colitis [54] ( anti-IL-25 neutralizing antibodies can inhibit oxazolone-induced colitis [55] ) . Thus , application of anti-IL-22 or anti-IL-25 antibodies to manipulate human disease would need to proceed with caution and be preceded by precise work-up of the immunological basis of the disease in the patient to be treated . The role of IL-25 has been assessed in TH2-mediated airways diseases as an early TH2-promoting factor [56–58] . In the context of TH1-mediated pathologies , IL-25 has been shown to suppress IL-17 and IFNγ production in infectious [26] and autoimmune diseases ( e . g experimental autoimmune encephalitis ( EAE ) [59] and diabetes [45] ) . Interleukin-25 has been found to inhibit the release of IL-1β , IL-12 ( p40 ) and TNFα from LPS-activated human CD14+ monocytes [60] which could in part explain its’ suppression of TH1-driven immunopathologies . Having found increased IL-25 expression in H . diminuta-infected IL-22-/- mice and that these mice were highly resistant to DNBS-induced colitis , a causal relationship between these two observations was tested via administration of IL-25 neutralizing antibodies [61] . First , the role of IL-25 during DNBS-colitis in IL-22-/- mice in the absence of H . diminuta infection was addressed . Consistent with the previous data , IL-22-/- mice displayed less severe DNBS-induced colitis compared to WT mice ( Fig 8 ) . However , IL-22-/- mice treated with DNBS and anti-IL-25 blocking antibodies had a severity of colitis that was macroscopically ( Fig 8A–8C ) and microscopically ( Fig 8D ) indistinguishable from WT mice that received DNBS only . Thus , in the absence of infection with H . diminuta ( an IL-25 trigger ) , IL-22 represses IL-25 during inflammatory responses induced by DNBS and when IL-25 is blocked the resistant phenotype observed in IL-22-/- mice is negated . In vivo immunoneutralizing of IL-25 in DNBS+H . diminuta-infected IL-22-/- mice resulted in a severity of colitis that was similar to DNBS-only treated mice , indicating a requirement for IL-25 in the anti-colitic effect evoked following infection with this helminth ( Fig 9 ) . These findings complement other studies in which IL-25 has been shown to down-regulate inflammatory gut disease: for example colitis induced in mice by bacterial peptidoglycan , TNBS , oxazolone or DSS [37 , 62 , 63] . Going forward it will be intriguing to test helminth therapy with/without IL-22 in chronic models of colitis and those driven by adaptive immunity such as the naïve T cell transfer model [4] . Assessment of the role of IL-22 in immunity and inflammation reveals that the impact of this cytokine is highly contextual , with convincing evidence in favour of anti- and pro-inflammatory roles [4 , 5 , 7 , 64] . While many of the functions of IL-22 in the gut promote protective anti-microbial responses , a pathogenic role for IL-22 has been described following infection with T . gondii [19] and Helicobacter pylori [65] . Less is known of the role of IL-22 in the host response to infection with helminth parasites . Increases in local IL-22 or IL-22+ cells have been described in response to gastrointestinal helminths [21 , 22] , yet the function of IL-22 was inferred not tested . The notable exception being the demonstration of impaired expulsion of nematodes in IL-22-/- mice that aligned with reduced goblet cell hyperplasia [20] . The role of IL-22 , if any , in regulating the response to cestode parasites has not hitherto been examined . Production of IL-22 can be evoked by IL-9 , IL-23 and microbial stimuli [3] and while IL-25 suppression of IL-22 has been shown [66] , less is known of the reciprocal interaction . We have found that increases in IL-25 mRNA in the parasitized intestine and IL-25 synthesis by enteric epithelia exposed to H . diminuta are suppressed by IL-22 . This is , to our knowledge , the first time IL-22 has been directly implicated in the control of helminth-evoked IL-25 , and complements earlier work showing that IL-22 inhibited IL-25 production by cytokine-treated murine airways epithelia [67] . Using the H . diminuta-mouse model system , data have been obtained that support the following conclusions: ( 1 ) absence of IL-22 reduces the early TH2 response to infection with helminth parasites , suggesting an important initial role for innate immunity against metazoan parasites; ( 2 ) IL-22 is an endogenous brake on helminth-provoked TH2 immunity , and in its’ absence there is heightened/prolonged local ( i . e . gut ) and systemic TH2 and immunoregulatory events ( e . g . IL-10 ) , likely driven in large part by the increase in IL-25; and , ( 3 ) by limiting the synthesis of IL-25 , IL-22 participates in the pathogenesis of DNBS-induced colitis and restricts the H . diminuta-suppression of colitis ( Fig 10 ) . Helminth therapy has been presented as a novel approach to auto-inflammatory disease [68] and we speculate that precise knowledge of the immunological basis of the disease would be important in selecting patients for helminth therapy .
Interleukin-22 deficient mice ( IL-22-/-: C57BL/6 background ) were bred at the University of Calgary ( pairs kindly provided by Dr . M . Kelly ( Univ . of Calgary ) ) . Mice were housed in a 12:12 hr light:dark cycle with free access to food and water and 8–9 weeks old male IL-22-/- and age-matched C57BL/6 control mice ( Charles River , QB , Canada ) were used throughout this study . As defined in the experiments , mice received 5 infective H . diminuta cysticercoids in 100 μl of sterile 0 . 9% NaCl by oral gavage and 8 days later colitis was induced [23] . In one experiment doses of 5 and 10 H . diminuta cysticercoids were compared . All experiments were conducted following the regulations specified by the Canadian Guidelines for Animal Welfare and were approved by the University of Calgary Health Science Animal Care Committee ( HSCCC ) with the protocol number AC13-0015 . At time-points post-infection , the small intestine was excised and flushed with 2 ml of 4°C PBS . The intestine was opened longitudinally and examined along with the flushed contents for H . diminuta . Colitis was induced by intrarectal ( ir . ) installition of 5 mg/mouse of DNBS ( MP Biomedicals Ohio , USA ) in 100 μl of 50% ethanol 3 cm into the colon . Weight was recorded daily for 3 days , the mice humanely necropsied and a macroscopic disease activity score on a 5 point scale based on weight loss , colon shortening , stool consistency and general appearance determined as previously [23] . A portion of mid-colon was excised , formalin fixed , paraffin embedded and 5 μm sections were collected on coded slided , stained with hematoxylin and eosin and a histopathology score determined on a 12-point scale [23] . The most distal 1 cm of colon was snap frozen in liquid nitrogen for myeloperoxidase ( MPO ) determination as measure of granulocyte , mainly neutrophil , infiltrate . MPO activity was determined by a kinetic assay in which H2O2 catabolism is measured , and 1 unit of MPO activity is the amount of enzyme required to degrade 1 μM of H2O2/min [23] . In other experiments , a 5 day exposure to 2 . 5% wt . /vol . DSS ( MW: 30 , 000–50 , 000; MP , Biomedicals , OH , USA ) was used to induce colitis . Mice were transferred to regular tap water on day 5 , and 3 days later were assessed for disease severity as described above . Formalin-fixed , paraffin-embedded mouse mid-small intestine was sectioned ( 5 μm ) , sections collected on coded slides and stained with periodic-acid Schiff’s stain to identify goblet cells [31] . Cells were counted on a per villus-crypt unit ( VCU ) basis , as defined by an intact , rounded villus tip and an even layer or enterocytes indicating lack of obligue sectioning . To identify mast cells , sections were deparaffinized followed by epitope retrieval with 10 mM sodium citrate buffer pH 6 . 0 . After washing sections were incubated in PE anti-mouse CD117 ( c-Kit ) antibody ( BioLegend , CA , USA ) ( 1:500 ) in blocking solution at 4°C overnight . Subsequently sections were washed in PBS , incubated in DAPI ( 0 . 1 μg/mL , 10 min . at room temperatura ) and after a final PBS wash , slides were mounted using ProLong Gold ( Cell Signaling Technology ) and examined with a Nikon 80i microscope and DXM1200C camera . Images were captured using NIS-Elements software ( Nikon ) , and representative images were processed in Adobe Photoshop ( Version 8 . 0 ) . At indicated times the spleen and mesenteric lymph nodes ( MLN ) were asceptically removed from WT and IL-22-/- H . diminuta-infected mice , cell suspensions generated and red blood cells lysed in ammonium chloride buffer [23] . Cells were adjusted to 3x106 /ml in RPMI 1640 medium supplemented with 10% FBS , 0 . 1 mM ( Gibco , USA ) . Cells were activated by treatment with concanavalin A ( 5 μg/ml ) and 48 hr later supernatants were collected and stored ( -80°C ) for cytokine measurements by ELISA . Interleukin ( IL ) -4 , IL-5 , IL-10 , IL-13 , IL-17 , IL-25 and IFNγ were measured by ELISA using paired antibodies and following the manufactures’ instructions ( R&D Systems Inc . , Minneapolis , USA ) . All samples were measured in duplicate and assays had dectection limits that ranged from 2–9 pg/ml . Spleens were aseptically excised and cell suspensions generated as above . Thereafter , 1x106 splenocytes were incubated with TrueStainX ( anti-CD16/32 ) for 10 min at 4°C and then stained . Cells were stained for 30 min with conjugated APC-CD4 ( Biolegend , San Diego , CA USA ) . After incubation with APC-CD4 antibody cells were washed in flow buffer ( PBS , 1%FBS and 0 . 1% NaN3 ) and intracellular staining for Foxp3 was performed following manufacturer’s protocol . Briefly , after surface staining cells were washed with flow cytometry buffer , then fixed and permeabilized with Foxp3 Fix/Perm and Foxp3 Perm buffers respectively . A final incubation with Foxp3-AlexaFluor 488 ( Biolegend , San Diego , CA ) was conducted for 30 min at room temperature in the dark . Data were acquired in a Attune cytometer and analyzed with Attune V . 6 . 1 software ( R&D systems ) . Small intestine was excised from non-infected and H . diminuta-infected WT and IL-22-/- mice , flushed with 4°C PBS , and the 3 cm portion of mid-intestine was cut in three pieces , placed in 1ml of TRizol Reagent ( Invitrogen , California , USA ) and homogenized for 60 seconds ( Polytron MR2100 , Kinematica AG , Switzerland ) . The RNA was extracted with chloroform/ethanol as previously [25] and 1 μg of RNA was used as the template for cDNA generation with the iScript DNA synthesis kit ( Bio-Rad , USA ) . Conditions for the PCR were denaturation 95°C for 2 min , 40 amplifying cycles of 95°C 15 sec , 55°C 15 sec , 68°C 20 sec and final temperature 4°C; primer sequences are presented in S10 Fig . At indicated times after H . diminuta infection ~1cm of jejunum was excised and homogenized in RIPA buffer ( 50 mM Tris-HCl , 150 mM NaCl , 1% NP-40 0 . 5% sodium deoxycholate and 0 . 1% SDS ) supplemented with protease inhibitor cocktail ( Promega , Madison Wisconsin USA ) . Protein concentration was determined by the Bradford assay ( Bio-Rad Laboratories Mississauga ON , Canada ) . Samples were normalized to 10 μg protein/μl and run by SDS-page ( 4% stacking , 8% separating ) and transferred to a nitrocellulose membrane . Membranes were blocked for 1 hr at room temperature in 5% skim milk in 0 . 1% TBS-tween buffer and then incubated overnight with purified anti-Foxp3 , 3 μg/ml ( Biolegend , California , USA ) . After washing , membranes were incubated with appropriate secondary antibody for 1 hr at room temperature and developed by exposing to western lightning plus enhanced chemiluminiscence solution ( PerkinElmer , Woodbridge ON , Canada ) for 1 min and using an automatic film developer . The mouse small intestinal epithelial cell line , IEC4 , was maintained by serial passage in DMEM medium supplemented with HEPES ( 1% ) , L-glutamine ( 10% ) , Pen/Strep ( 1% ) and FBS ( 5% ) ( all from Gibco , USA ) . One-million IEC4 cells were seeded in 6-well plates and cultured for 48 hr . Scolices and 2 cm of strobila of H . diminuta retrieved from the small intestine of rats or IL-4 receptor-α-/- mice ( fail to expel H . diminuta ) were exposed to a cocktail of antibiotics ( Gentamicin solution , Sigma , St . Louis , Mo , USA ) for 2 hr . A single worm was added to IEC4 monolayers ± recombinant IL-22 ( 5 ng/ml; Biolegend , CA , USA ) , and supernatants collected for measurement of IL-25 and then total RNA extracted . To determine the role of IL-25 , IL-22-/- were treated with a single ip . injection of 100 μg of an anti-IL-25 blocking antibody ( clone 35B , Biolegend , CA , USA ) ~10 min prior to DNBS ir . delivery and the severity of colon inflammation was assessed 72 hr later ( as above [23] ) . Following a protocol to transfer colonic microbiota between mice [51 , 69] , WT and IL-22-/- mice were transferred to cages with fresh bedding and 24 hr later mice were swapped into the opposing strains cage without a bedding change for 24 hr ( coprophagy allows blending of the microbiota between the two strains ) . This cycle of swapping mice between cages was continued for 2 weeks . On day one of the procedure all mice were treated with kanamycin ( 40 mg/kg ) , gentamicin ( 3 . 5 mg/kg ) , colistin ( 4 . 2 mg/kg ) , metronidazole ( 21 . 5 mg/kg ) for 3 days in their drinking water followed by an ip . injection of vancomycin ( 4 . 5 mg/kg ) . On day 15 mice were anesthetized and given DNBS ( 5 mg/mouse ) intrarectally and colitis severity was assessed 72 hr later . Bone marrow was flushed from the long bones of the legs of WT and IL-22-/- mice via a sterile 27 gauge needle , the red blood cells lysed and the cells were incubated in RPMI 1640 medium ( Gibco , USA ) supplemented with 20% FBS , HEPES , Glutamax and antibiotic ( Penicillin-streptomycin Sigma , St . Louis , Mo , USA ) for 7 days in presence of 20 ng/ml murine M-CSF . On days 2 and 4 cells were treated with fresh medium containing macrophage-colony stimulating factor ( M-CSF ) . At day 7 , mature macrophages were harvested and seeded at 2 . 5x105 in 24-well plates in above-mentioned medium and incubated with PBS-soluble crude H . diminuta antigen ( HdAg: 100 μg/ml [70] ) for 24 hr . As additional control , macrophages were also stimulated with LPS ( 10–1000 ng/ml ( Sigma , St . Louis , MO , USA ) ) . Supernatants were collected and assayed for TNFα by ELISA . Data are presented as mean ± the standard error of the mean ( SEM ) and statistical differences were determined by one-way ANOVA followed by post-hoc analysis with Student’s t test or Kneuman’s Keuls test and p<0 . 05 accepted as a statistically significant difference ( Graph Pad prism V5 software , La Jolla , CA , USA ) . | Interleukin ( IL ) -22 , produced by innate and adaptive immune cells , plays a complex role in immunity; under specific conditions , targeting this cytokine could treat inflammatory diseases . The hygiene hypothesis suggests infection with helminth parasites could ameliorate inflammation . Here we show that IL-22 is required to activate early events ( i . e . type 2 cytokines and mucin expression ) in the response to the non-invasive cestode Hymenolepis diminuta . Strikingly , expression of regulatory factors ( IL-10 , IL-25 , Foxp3 ) , which arise following H . diminuta infection , were either enhanced or sustained in IL-22-/- mice , uncovering a novel role for IL-22 as a brake for these regulatory events following infection with this parasitic helminth . Moreover , DNBS-induced colitis was significantly less severe in IL-22-/- compared to wild-type mice: IL-22-/- mice infected with H . dimunta 8-days prior to the induction of colitis had negligible disease . Immunoneutralization of IL-25 exaggerated DNBS-induced colitis in the IL-22-/- mice and ablated the anti-colitic effect of infection with H . diminta . Thus , while immune events in the early response to infection with H . diminuta are delayed in IL-22-/- mice ( as is worm expulsion ) , the compensatory enhancement of IL-25 ( and other immunoregulatory elements ( e . g . IL-10 ) ) provide resistance to colitis and also promote the anti-colitic effect driven as a consequence of the response to infection with H . diminuta . The data confirm the complex role of IL-22 in intestinal immunity . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] | [
"invertebrates",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"cytokines",
"helminths",
"immunology",
"parasitic",
"diseases",
"animals",
"nematode",
"infections",
"colitis",
"developmental",
"biology",
"gastroenterology",
... | 2016 | IL-22 Restrains Tapeworm-Mediated Protection against Experimental Colitis via Regulation of IL-25 Expression |
Treponema pallidum infection evokes vigorous immune responses , resulting in tissue damage . Several studies have demonstrated that IL-17 may be involved in the pathogenesis of syphilis . However , the role of Th17 response in neurosyphilis remains unclear . In this study , Th17 in peripheral blood from 103 neurosyphilis patients , 69 syphilis patients without neurological involvement , and 70 healthy donors were analyzed by flow cytometry . The level of IL-17 in cerebrospinal fluid ( CSF ) was quantified by ELISA . One-year follow up for 44 neurosyphilis patients was further monitored to investigate the role of Th17/IL-17 in neurosyphilis . We found that the frequency of Th17 cells was significantly increased in peripheral blood of patients with neurosyphilis , in comparison to healthy donors . IL-17 in CSF were detected from 55 . 3% neurosyphilis patients ( in average of 2 . 29 ( 0–59 . 83 ) pg/ml ) , especially in those with symptomatic neurosyphilis ( 61 . 9% ) . CSF IL-17 was predominantly derived from Th17 cells in neurosyphilis patients . Levels of IL-17 in CSF of neurosyphilis patients were positively associated with total CSF protein levels and CSF VDRL ( Venereal Disease Research Laboratory ) titers . Notably , neurosyphilis patients with undetectable CSF IL-17 were more likely to confer to CSF VDRL negative after treatment . These findings indicate that Th17 response may be involved in central nervous system damage and associated with clinical symptoms in neurosyphilis patients . Th17/IL-17 may be used as an alternative surrogate marker for assessing the efficacy of clinical treatment of neurosyphilis patients .
Syphilis , a sexually transmitted multi-stage disease caused by the spirochete Treponema pallidum , remains to be a global public health problem with an estimated 12 million new cases annually [1] . In recent years , China has experienced a resurgence of syphilis cases , with the national incidence rate of 32 . 04 per 100 , 000 population and with 429 , 677 new cases reported in 2011 [2] . T . pallidum invades the human host through genital or oral mucosa , abraded skin , enters lymphatic system and bloodstream , and then disseminates to different organs . Without treatment , this spirochetal pathogen is able to survive in the human host for several decades , causing damage in multiple organs including nervous system ( neurosyphilis ) [3] , [4] . Neurosyphilis is a frequent and protean clinical manifestations ranging from headache and oculopathy to more serious conditions such as cerebrovascular events , paretic and tabes dorsalis [5] . The mechanisms underlying the development of neurosyphilis remain poorly understood . T . pallidum can invade the CNS at any stage of infection and provokes robust cellular immune response [6] . In the non-human primate models , strong T helper ( Th ) 1-type immune response can contribute to the clearance of T . pallidum in CNS [6] . The immune response elicited during infection , although aimed to eliminate organisms , may also contribute to the pathogenesis . Cytokines produced by T lymphocytes are critical for regulation of both protective and pathogenic immune responses [7] . Th17 cells , with the hallmark of producing cytokine IL-17 , were identified as a subset of CD4+ T helper cells . Emerging evidence has demonstrated that Th17 cells contribute to clearance of diverse organisms ( Mycobacterium tuberculosis , Pneumocystis carinii , Candida albicans and Klebsiella pneumonia et al . ) [8] , [9] , [10] , [11] . On the other hand , Th17 also mediates strong immunopathology in chronic infection . Anti-IL-17 and anti-IL-17R treatments could prevent severe Borrelia-induced destructive arthritis [12] . Hence , Th17 response in infection may be involved in both protection and progression/chronic infection . Previous studies reported an increase of IL-17 in secondary syphilitic lesion and peripheral blood [13] , [14] . Recently , Pastuszczak et al . also showed elevated CSF IL-17 levels in early asymptomatic neurosyphilis [15] , suggesting that IL-17 may be involved in local immune response to T . pallidum infection . In this study , we performed a comparative analysis of Th17/IL-17 in peripheral blood and CSF in asymptomatic and symptomatic neurosyphilis patients , and evaluated the relationship between CSF IL-17 level and the clinical outcomes . Our results suggested that Th17/IL-17 is a contributing factor to the immunopathology of neurosyphilis , and may be used to monitor the prognosis of treatments of syphilis infected patients .
This study was performed at the Shanghai Skin Disease Hospital between Aug . 2010 and Dec . 2012 . The study was approved by the Ethics Committee of the Shanghai Skin Disease Hospital . Written informed consents were obtained from all participants . Patients were identified and referred for enrollment by dermatologists , neurologists , psychiatrists and ophthalmologists after careful examination and evaluation . Syphilis was diagnosed at each stage of infection by a combination of compatible history , clinical features and the results of nontreponemal and treponemal tests of serum and CSF samples . The exclusion criteria include positive HIV infection; prior history of syphilis infection , or history of syphilis treatment ( except for 7 serofast patients ) ; history of systemic inflammation , autoimmune disease , other underlying acute or chronic disease , receiving anti-inflammatory medications , immunocompromised conditions , or use of antibiotics or immunosuppressive medications in the last four weeks . 70 healthy donors , who visited Shanghai Skin Disease Hospital voluntarily for a medical check-up for the purpose of STD prevention , were recruited to the study . All healthy subjects were negative for HIV and serological tests for syphilis ( i . e . , both serum RPR and TPPA negative ) , and did not have any clinical symptoms consistent with T . pallidum infection . In this study , three groups of patients were included: 1 ) neurosyphilis group ( including 40 subjects with asymptomatic neurosyphilis , 4 with meningovasculitis , 39 with general paresis , 8 with tabes dorsalis , and 12 with ocular neurosyphilis ) ; 2 ) non-neurosyphilis group with normal CSF WBC count , CSF protein concentration and CSF-VDRL negative ( including 13 subjects with primary syphilis , 30 with secondary syphilis , 19 with latent syphilis , and 7 with serofast syphilis ) ; 3 ) 70 healthy donors . In this study , the serofast state must be met the following three criteria: i ) syphilitic patients , despite receiving recommended standard treatment ( according to Chinese National STI Treatment Guidelines ) , whose nontreponemal test titers ( RPR ) persists positive for at least two years of follow-up evaluation . ii ) patients who denied high risk sexual behaviors ( re-infection ) following treatments; and iii ) patients with RPR titers declined fourfold within 6 months after therapy . Peripheral blood from healthy donors was used for peripheral blood mononuclear cells ( PBMC ) isolation and for measurement of the baseline of the levels of IL-17+ cells and the frequency of Th17 cells . Since it is difficult to collect CSF from healthy donors , we used a separate control group of 29 patients who underwent orthopaedic or stone surgery ( gall stone , kidney stone ) but were serum RPR and TPPA negative , whose CSF samples were collected prior to spinal anaesthesia . The baseline level of IL-17 in CSF was determined using samples from the control group . All groups were well matched in the categories of gender and age . Additional information on the patient groups were presented in Table 1 . CSF samples were stored at −70°C and thawed once before analyses . First , all neurosyphilis patients have positive serum RPR and TPPA tests . The diagnosis of confirmed neurosyphilis includes reactive CSF-VDRL ( Venereal Disease Research Laboratory ) and CSF-TPPA tests in the absence of substantial contamination of CSF with blood . Presumptive neurosyphilis was defined as a nonreactive CSF-VDRL test but reactive CSF-TPPA with either or both of the following: ( i ) CSF protein concentration >45 mg/dl or CSF white blood cell ( WBC ) counts≥8/µl in the absence of other known causes for the abnormalities; ( ii ) clinical neurological or psychiatric manifestations consistent with neurosyphilis without other known causes for such abnormalities [16] , [17] . Neurosyphilis is categorized as: asymptomatic , meningovascular , paretic , ocular and tabetic neurosyphilis . Asymptomatic neurosyphilis is defined by the presence of CSF abnormalities consistent with neurosyphilis and the absence of neurological/psychiatrical symptoms or signs . Meningovasculitis is defined by clinical features of meningitis and magnetic resonance image ( MRI ) evidence of brain lesions and/or a stroke syndrome . General paresis is characterized by personality changes , dementia and psychiatric symptoms including mania or psychosis . Patients with sensory loss , ataxia , lancinating pains , bowel and bladder dysfunction were considered as Tabes dorsalis . Ocular neurosyphilis ( those who had ocular signs or symptoms but with normal CSF index were not included in this study ) is defined by the presence of CSF abnormalities consistent with neurosyphilis and ocular signs or symptoms ( worsening visual acuity and visual fields , floaters , papillitis , uveitis ) . All these forms of neurosyphilis should have no other known causes for these clinical abnormalities . A complete list of information of neurosyphilis patients were listed in Table 2 . According to Chinese National STI treatment Guidelines , syphilis patients without CNS involvement were treated with benzathine penicillin 2 . 4MU/qw intramuscular for 1 or 2 weeks for early syphilis , and 3 weeks for late or unknown duration syphilis . If allergic to penicillin , ceftriaxone 0 . 25 g/day intramuscular for 10 days were given . Neurosyphilis patients were given aqueous crystalline pencillin G , 4MU intravenously every 4 h for 14 days , or ceftriaxone intravenously with 2 g daily for 10 days if allergic to penicillin . In the 103 neurosyphilis patients , 80 patients were treated with aqueous crystalline pencillin G , 4MU intravenously every 4 h for 14 days , 23 patients were treated with ceftriaxone intravenously with 2 g daily for 10 days . All patients were asked for follow-up after treatment . Patients were selected if the patient's written informed consent was obtained . The exclusion criteria include positive HIV infection; history of syphilis or syphilis treatment; history of systemic inflammatory , autoimmune disease , other underlying acute or chronic disease , patients receiving anti-inflammatory medications , immunocompromised , or using antibiotics or immunosuppressive medications in the last four weeks . In this study , 44 neurosyphilis patients were enrolled and followed up at Shanghai Skin Disease Hospital . Patients returned for follow-up visits at 3 , 6 , 9 and 12 months after treatment . All patients underwent lumbar puncture at the 3-month visit , and lumbar punctures were repeated at 6 , 9 and 12 months if the previous CSF profile was abnormal . Blood samples were collected at each follow-up visits . All patients completed their 12 months follow-up visit . For the CSF-VDRL and the serum RPR test , a 4-fold decrease or more in titer or reversion to a nonreactive result was defined as a normal response . Stepwise Cox regression models were used to determine the influence of the following factors on the likelihood of normalization of each measure and the improvement of clinical symptoms: ( 1 ) neurosyphilis treatment regimen ( intravenous ceftriaxone , vs . intravenous aqueous penicillin G ) ; ( 2 ) syphilis stage ( secondary and early latent vs . late latent and syphilis of unknown duration ) ; ( 3 ) baseline laboratory values ( greater or less than the median value for those subjects with each abnormality ) ; ( 4 ) CSF IL-17 levels: CSF IL-17 negative ( <0 . 5 pg/ml ) and CSF IL-17 positive ( ≥0 . 5 pg/ml ) ; and ( 6 ) clinical symptoms . Whole blood samples ( 5 ml ) from syphilis patients and healthy donors were used for peripheral blood mononuclear cells ( PBMC ) isolation . PBMC were purified from peripheral blood by centrifugation using a Ficoll-Hypaque gradient ( Axis-Shield ) . Because resting cells do not normally produce cytokines , cells were stimulated in vitro in order for the respective cytokine genes to be activated for intracellular cytokine staining . Phorbol myristate acetate ( PMA ) and ionomycin are unspecific stimulator that trigger a strong production of cytokines in vitro and are widely used to evaluate intracellular cytokine production from various T lymphocyte subpopulations [18] . Monensin is used to prevent the intracellular transport of cytokines from Golgi apparatus for enhancing the sensitivity of the detection . Accordingly , PBMC were seeded into 24-well culture plates ( Corning ) at 2×106 cells/well and stimulated ex vivo with PMA ( 50 ng/ml ) ( Sigma ) and ionomycin ( 1 µg/ml ) ( Sigma ) for 4 hours . Monensin ( 2 uM ) ( eBioscience ) was then added at the start of stimulation . CSF ( 10 ml ) was centrifuged at 4°C immediately after spinal tap , and cells were stimulated as described above . For intracellular staining , cells were first stained with ECD-labeled anti-human CD3 ( Clone UCHT1 , Beckman ) , FITC-labeled anti-human CD4 ( Clone RPA-T4 , Biolegend ) and then fixed and permeabilized using Perm/Fix solution ( Biolegend ) at room temperature for 20 minutes . Cells were washed with Perm/Wash buffer ( Biolegend ) and stained with PE-labeled anti-human IL-17A ( Clone BL168 , Biolegend ) . Mouse IgG1 and IgG2 ( BD Biosciences ) were used as isotype controls . After staining , cells were analyzed with Epics XL ( Beckman Coulter ) and FlowJo software ( Tree Star ) . Lymphocytes were gated according to forward and side scatter characteristics and CD4+T cells were gated based on CD3 and CD4 expression . IL-17 positive lymphocytes , CD3+ , CD4+ T cells were defined by setting regions with the lower limits for cytokine positivity determined from isotype antibody . IL-17 levels in CSF were determined using human IL-17 Quantikine ELISA kits ( eBioscience ) according to manufacturer's instruction . The sensitivity for detecting IL-17 is 0 . 5 pg/ml . Data were presented as median and range ( min , max ) . Differences between the groups were analyzed using the nonparametric Mann-Whitney U test . The detection rates between the groups were assessed using χ2 test or Fisher's exact test . Spearman correlation analysis was performed between the levels of IL-17 and other parameters . Stepwise Cox regression models were used to determine the influence of the factors on the likelihood of normalization of each measure . All statistical analyses were performed using SPSS 17 . 0 software . A value of P<0 . 05 was considered significant .
To investigate the potential role of Th17 in neurosyphilis , we first examined the frequency of IL-17+ among lymphocytes , CD3+ , CD4+ T populations in PBMC . The baseline frequency of total IL-17+ cells , and IL-17+ CD3+cells , and IL-17+ CD4+ T cells ( Th17 ) of PBMC in healthy individuals were 0 . 86% ( 0 . 19%–1 . 58% ) , 1 . 33% ( 0 . 48%–3 . 2% ) and 1 . 7% ( 0 . 56%–2 . 76% ) , respectively ( Fig . 1A , 1B & 1C ) . We observed a significant increase in the frequencies of IL17+ , CD3+IL-17+ and Th17 cells in syphilis patients with either non-neurosyphilis or neurosyphilis compared to those in healthy individuals ( Fig . 1A , 1B & 1C ) . However , there was no significant difference in the frequencies of IL-17+ cells , CD3+IL-17+ and Th17 cells in PBMC between syphilis patients without neurological involvement ( including primary , secondary , latent and serofast syphilis patients ) and neurosyphilis patients ( Fig . 1A , 1B & 1C ) . To further investigate whether Th17 cells in peripheral blood are different between diverse clinical presentations of neurosyphilis , we divided neurosyphilis patients into two groups , asymptomatic ( n = 40 ) and symptomatic neurosyphilis patients ( n = 63 ) . We then compared the Th17 cell frequency in PBMC between these two groups . As shown in Fig . 1D , 1E & 1F , patients with symptomatic neurosyphilis had significant higher percentage of total IL-17+cells , CD3+IL-17+ and Th17 in PBMC than the patient group with asymptomatic neurosyphilis . Since neurosyphilis patients had increased levels of Th17 cells in peripheral blood , we further investigated the IL-17 levels in CSF of these patients . We first compared the detection rate of IL-17 in CSF between neurosyphilis patients and non-neurosyphilis patients . We found that there was five-fold higher detection rate of IL-17 in CSF in neurosyphilis patients than that in non-neurosyphilis patients ( Fig . 2A ) . The average levels of CSF IL-17 was also significantly higher in neurosyphilis patients ( 2 . 29 pg/ml ) ( range of 0–59 . 83 pg/ml ) than that in non-neurosyphilis patients ( 0 pg/ml ) ( range of 0–2 . 60 pg/ml ) ( Fig . 2B ) . IL-17 was not detected in CSF of the control group ( Fig . 2A & B ) . We further compared the levels of CSF IL-17 between patients with asymptomatic and symptomatic neurosyphilis . The detection rates of CSF IL-17 were 47 . 5% and 61 . 9% in asymptomatic and symptomatic neurosyphilis , respectively . The level of CSF IL-17 in symptomatic neurosyphilis patients ( 4 . 91 pg/ml , range from 0 to 59 . 83 pg/ml ) was significantly higher than that in asymptomatic neurosyphilis patients ( 0 . 715 pg/ml , range from 0 to 44 . 27 pg/ml ) . Noted that the symptomatic neurosyphilis patient group included meningovascular , paretic , ocular and tabetic neurosyphilis . Further examination showed that the level of CSF IL-17 was the highest among paretic patients ( 7 . 6 pg/ml , range from 0 to 38 . 07 pg/ml ) ( Table 3 ) . T . pallidum is capable of invading central nervous system and damaging local tissues . There are detectable CSF abnormalities in neurosyphilis patients , including positive CSF VDRL , pleocytosis , and/or elevated protein concentration [19] . These measurements correlate well with the disease activity [19] . Since the above data showed that neurosyphilis patients had increased CSF IL-17 levels , we further investigated a possible relationship between CSF IL-17 levels and other measurements . As shown in Fig . 3 , there was a significant positive correlation between CSF IL-17 levels and CSF protein concentrations or CSF VDRL titer , but not with the CSF WBC counts . In some neurosyphilis patients , CSF IL-17 was not detected . We thus further investigated whether there are certain factors which may contribute to this phenomenon . We found that there were no differences in age , sex , the baseline serum RPR titer , or duration of symptoms prior to diagnosis between the IL-17 positive and IL-17 negative neurosyphilis groups . However , the IL-17 positive group had higher CSF protein concentration and CSF VDRL titer and higher frequency of symptomatic neurosyphilis than that of the IL-17 negative group ( Table 4 ) . We further analyzed IL-17-producing cells in CSF of neurosyphilis patients . Because of the limited sample sizes and lymphocyte cell numbers in CSF collected from neurosyphilis patients , CSF cells from 14 neurosyphilis patients who had high levels of CSF pleocytosis ( >50 cells/ul ) were collected and stimulated for intracellular staining for the purpose of this study . 14 subjects included 6 ( 42 . 9% ) subjects with asymptomatic neurosyphilis , 5 ( 35 . 7% ) subjects with paretic , 2 ( 14 . 3% ) subjects with ocular neurosyphilis , 1 ( 7 . 14% ) subjects with meningovascular neurosyphilis . The average percentage of CD4+ T cells was 58 . 28% ( 51 . 65%–80 . 1% ) of total lymphocytes , and Th17 ( CD3+CD4+IL-17+ cells ) was 1 . 8% ( 0 . 25%–4 . 6% ) ( Fig . 4 ) . However , Th17 cells accounted for 88 . 8% ( 45 . 1%–100% ) of total IL-17+ cells ( Fig . 4 ) , indicating that Th17 is the dominant IL-17-producing cells and may play an important role in neurosyphilis . Among 44 subjects with confirmed neurosyphilis , 22 ( 50% ) subjects were asymptomatic neurosyphilis , 15 ( 34 . 1% ) were paretic neurosyphilis , 5 ( 11 . 4% ) were ocular neurosyphilis , 2 ( 4 . 5% ) were meningovascular neurosyphilis . All enrolled patients were routinely under followed-up examination and treated with standard therapy according to the Chinese treatment Guidelines . Factors that were included in the final regression models of normalization of each laboratory measure , the improvement of clinical symptoms and their HRs are shown in Table 5 . Factors that may improve laboratory markers or clinical symptoms were analyzed using the final regression model ( Table 5 ) . The neurosyphilis treatment regimen did not influence normalization of any of the 4 laboratory markers and the improvement of clinical symptoms . Normalization of the CSF protein concentration was more likely to occur in subjects with early syphilis ( p = 0 . 018 ) . CSF-VDRL reactivity was less likely to become normal in patients with positive CSF IL-17 ( p = 0 . 04 ) and with higher baseline CSF-VDRL titer ( p = 0 . 019 ) . Serum RPR reactivity was more likely to return to normal in subjects with higher baseline serum RPR titers ( p = 0 . 008 ) .
T . pallidum remains one of the human pathogens that cannot be cultivated in vitro to-date . A suitable animal model for studying the pathogenesis of syphilis is also lacking . These obstacles have greatly hindered the effort of elucidating the basic immunobiological traits of syphilis . As a consequence , little is known about how T . pallidum causes damage to the central nervous system . IL-17 , a potent proinflammatory cytokine , plays a key role in the induction and development of tissue injury . IL-17 results in an increased production of ICAM-1 , IL-6 and IL-8 , and an increased synergy of many effects of IL-1β and TNF-α , which enhances the local inflammation and leads to inflammatory destruction [20] , [21] . In this study , we observed an elevated CSF IL-17 in neurosyphilis patients . A similar scenario has been observed in infectious and auto-immune CNS disorders [22] , [23] . Furthermore , we found that the level of CSF IL-17 is positively associated with CSF VDRL titer and total CSF protein in neurosyphilis patients . These findings suggest that IL-17 may involve in the CNS damage in neurosyphilis patients . Syphilis is known as a “great imitator” because it is protean , especially neurosyphilis . Based on the patient's clinical and laboratory features , neurosyphilis is divided into five diagnostic categories , including asymptomatic , meningitis , meningovascular , paretic , and tabetic neurosyphilis [5] . Meningitis involves diffuse inflammation of the meninges with signs and symptoms of meningitis including headache , photophobia , nausea , vomiting , cranial nerve palsies , and occasionally seizures . It is diagnosed within 12 months after T . pallidum infection but it is relatively rare [5] . Unfortunately , no meningitis neurosyphilis patients enrolled in this study , and thus , the involvement of IL-17 in this stage of neurosyphilis remains unknown . The pathogenic mechanisms underlying different clinical presentations of neurosyphilis are largely unknown . In this study , we found that higher levels of IL-17 were observed in CSF of symptomatic neurosyphilis patients , especially in paretic patients , compared with asymptomatic neurosyphilis patients . Moreover , the higher CSF protein and VDRL titer were observed in symptomatic neurosyphilis patients . These results suggest that IL-17 may be associated with clinical symptoms in neurosyphilis patients . Asymptomatic neurosyphilis does occur in both early and latent stages of syphilis . It is reported that there was an elevated CSF IL-17 level in early asymptomatic neurosyphilis patients , which correlated with the extent of CSF abnormality [15] . In our study , CSF IL-17 could be detected in 66 . 7% ( 12/18 ) of early asymptomatic neurosyphilis patients . It is believed that asymptomatic neurosyphilis patients may develop to late neurological complications [24] . Moreover , the extent of abnormalities of CSF positively correlated with the probability of developing late neurological complications [25] . Based on these notions , we suggested that some early asymptomatic neurosyphilis patients might have persistent IL-17 inflammation response , which could damage the CNS , resulting in neurological symptoms . Regrettably , there has been no study to compare long-term outcomes between CSF IL-17 negative and positive asymptomatic neurosyphilis patients . Pastuszczak et al . identified that there was a strong correlation between CSF IL-17 and CSF pleocytosis in early asymptomatic neurosyphilis patients [15] . But our results indicated that there was no correlated between the level of CSF IL-17 and CSF pleocytosis . The CSF pleocytosis in neurosyphilis was related to the syphilis stage besides to the CNS damage . There was a marked pleocytosis in patients with acute meningeal syphilis . In late stage , CSF WBC counts in some neurosyphilis patients were less or even normal and were inconsistent with clinical symptoms . In up to 10% of patients with tabes referred to as the “burned out” stage , the CSF cell count may be normal [5] . In our study , there were early and late stage neurosyphilis patients . There were some paresis patients the CSF WBC counts were normal , though the clinical manifestations were severe . Therefore , according to the data , the CSF WBC counts were not always correlated with the degree of CNS damage . The different stage patients were enrolled in our study , leading to be inconsistent with the previous results . Besides CD4+ T cells , other cells are capable of secreting IL-17 [26] . It was previously shown that IL-17 is mainly secreted by CD4+ T cells ( Th17 ) in CSF in patients with chronic inflammatory demyelinating polyradiculoneuropathy ( CIDP ) [27] . In this study , we observed the CD4+ T cells were accumulated in CSF in neurosyphilis patients , and they were the dominant IL-17-producing cells . This finding suggests that Th17 response is a part of the local CNS response in a sub-population of neurosyphilis patients . Our results showed that Th17 cells were increased in CSF of neurosyphilis patients . The mechanisms underlying the increase of Th17 in CNS remain unclear . IL-17 can disrupt the tight junction molecules and activates the endothelial contractile machinery , leading to disruption of blood brain barrier ( BBB ) [28] . Thus , Th17 in CSF may be a consequence of passive diffusion from blood . On the other hand , microbial lipopeptides such as Helicobacter pylori HP-NAP and B . burgdorferi NapA , can induce Th17 differentiation and production of IL-17 [29] , [30] . In this regard , T . pallidum TpF1 is a protein homolog of HP-NAP and NapA [31] , which may be capable of promoting Th17 differentiation and expansion in CNS . Interestingly , recent data showed that TpF1 can stimulate Treg cell differentiation [32] . The mechanisms underlying the increase of Th17 in CNS in neurosyphilis need to be further elucidated . Because Th17 response may induce the immune-mediated CNS injury , we further evaluated the relationship between IL-17 and the clinical outcome of neurosyphilis . The baseline CSF-VDRL titer , and serum RPR titer influenced the likelihood of normalization of each parameter , consistent with previous studies [33] . However , we observed that CSF IL-17 positive neurosyphilis patients were 2 . 43 times less likely to normalize CSF-VDRL reactivity , even after taking into account the baseline CSF-VDRL titer and the stage of syphilis . CSF VDRL titer may reflect the T . pallidum burden and the extent of CNS damage . T . pallidum invaded CNS can induce Th17 immune response and CSF IL-17 were positively correlated with CSF VDRL titer . So positive CSF IL-17 in patients may reflect higher number of T . pallidum spirochetes in CSF . Since longer time would be required to clear a higher number of T . pallidum burden , the likelihood of normalization of CSF VDRL reactivity at the end of the observation would be lower . Because the sample size is limited ( the number of subjects included in the regression analyses was only ranged from 21 to 44 patients in this study ) , a large sample study is needed to further understanding the true immune response at different stages of neurosyphilis and its clinical significance . In conclusion , our findings demonstrate that neurological damage in syphilis patients is associated with increased CSF Th17/IL-17 response . CSF IL-17 may be used to evaluate the clinical outcome of treatment of neurosyphilis . | Syphilis , caused by the bacterium Treponema pallidum , can progress to affect the central nervous system ( CNS ) and cause damage in the brain and spinal cord , which is called neurosyphilis . While many neurosyphilis patients may not have any symptoms , some patients develop severe symptoms which can be life-threatening . Th17 cells are a subset of CD4+ T helper cells producing the hallmark cytokine IL-17 , which are essential for effective antimicrobial host defense and are also involved in tissue damage . In this study we conduct a comparative analysis of Th17/IL-17 in the peripheral blood and cerebrospinal fluid ( CSF ) of syphilis patients without neurological abnormalities , and neurosyphilis patients with or without symptoms . Th17 frequency in peripheral blood was significantly increased in neurosyphilis . CSF IL-17 was increased in neurosyphilis patients , especially in those with symptomatic neurosyphilis . Levels of CSF IL-17 in neurosyphilis patients were positively associated with CNS damage . Notably , neurosyphilis patients with undetectable CSF IL-17 had better outcome upon treatment . These findings indicate that the Th17 response may be involved in central nervous system damage , clinical symptoms and prognosis of treatment of neurosyphilis patients . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"bacterial",
"diseases",
"infectious",
"diseases",
"infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"medicine",
"and",
"health",
"sciences",
"neurology"
] | 2014 | Increased Interleukin-17 in Peripheral Blood and Cerebrospinal Fluid of Neurosyphilis Patients |
Cerebral malaria claims the lives of over 600 , 000 African children every year . To better understand the pathogenesis of this devastating disease , we compared the cellular dynamics in the cortical microvasculature between two infection models , Plasmodium berghei ANKA ( PbA ) infected CBA/CaJ mice , which develop experimental cerebral malaria ( ECM ) , and P . yoelii 17XL ( PyXL ) infected mice , which succumb to malarial hyperparasitemia without neurological impairment . Using a combination of intravital imaging and flow cytometry , we show that significantly more CD8+ T cells , neutrophils , and macrophages are recruited to postcapillary venules during ECM compared to hyperparasitemia . ECM correlated with ICAM-1 upregulation on macrophages , while vascular endothelia upregulated ICAM-1 during ECM and hyperparasitemia . The arrest of large numbers of leukocytes in postcapillary and larger venules caused microrheological alterations that significantly restricted the venous blood flow . Treatment with FTY720 , which inhibits vascular leakage , neurological signs , and death from ECM , prevented the recruitment of a subpopulation of CD45hi CD8+ T cells , ICAM-1+ macrophages , and neutrophils to postcapillary venules . FTY720 had no effect on the ECM-associated expression of the pattern recognition receptor CD14 in postcapillary venules suggesting that endothelial activation is insufficient to cause vascular pathology . Expression of the endothelial tight junction proteins claudin-5 , occludin , and ZO-1 in the cerebral cortex and cerebellum of PbA-infected mice with ECM was unaltered compared to FTY720-treated PbA-infected mice or PyXL-infected mice with hyperparasitemia . Thus , blood brain barrier opening does not involve endothelial injury and is likely reversible , consistent with the rapid recovery of many patients with CM . We conclude that the ECM-associated recruitment of large numbers of activated leukocytes , in particular CD8+ T cells and ICAM+ macrophages , causes a severe restriction in the venous blood efflux from the brain , which exacerbates the vasogenic edema and increases the intracranial pressure . Thus , death from ECM could potentially occur as a consequence of intracranial hypertension .
Plasmodium falciparum is responsible for an estimated 600 , 000 deaths annually , principally in children under the age of five [1] . Clinical symptoms range from intermittent fevers and chills to potentially fatal complications including severe anemia and cerebral malaria [2] . The mortality rate in comatose pediatric patients , most frequently due to respiratory arrest , is 15–20% despite optimal medical care [3] , but the underlying pathology is unclear . Molecular and cellular mechanisms involved in the pathogenesis of human cerebral malaria ( HCM ) include a predominantly pro-inflammatory cytokine profile , endothelial activation via the NF-κB pathway with upregulation of adhesion molecules , glia cell activation , and sequestration of infected red blood cells ( iRBC ) , monocytes , and platelets within brain capillaries [3]–[6] . However , the cellular mechanisms associated with HCM cannot be directly observed in the human brain . Ophthalmological examination of the retinal pathology generally correlates with course and etiology of malarial encephalopathy [2] , [7] , but despite significant recent improvements [8] , this technique lacks the resolution to observe the dynamic behavior of individual iRBC , leukocytes , and platelets , their exact location within the microvasculature , mechanisms of vascular leakage or possibly occlusion , and the sequence of these events . Elucidation of CM pathogenesis therefore requires the use of a robust small animal model that closely reflects clinical symptoms , histopathology , and immune mechanisms associated with the pathophysiology of HCM . P . berghei ANKA ( PbA ) infected CBA , Swiss Webster , or CB57Bl/6 mice represent a well-characterized and widely used model for experimental cerebral malaria ( ECM ) that shares a number of similarities with P . falciparum HCM [5] , [6] , [9]–[12] . Both ECM and HCM are characterized by severe vasculopathy , i . e . endothelial activation and dysfunction with increased expression of adhesion molecules such as ICAM-1 , VCAM-1 , and E-selectin , upregulation of inflammatory cytokines , reduced blood flow , vascular leakage , acute edema of both vasogenic and cytotoxic origin , and microhemorrhages leading to neurological impairment [12]–[17] . Platelet activation , dysregulation of the coagulation cascade , thrombocytopenia , and platelet accumulation in the brain are also found in both HCM and ECM [18]–[20] . We have previously shown by intravital microscopy ( IVM ) that platelet marginalization and blood brain barrier ( BBB ) disruption are central to ECM pathophysiology [21] . Platelets are thought to impair vascular repair and increase BBB permeability by potentiating the iRBC-induced endothelial damage in the early stages of HCM development [22]–[24] . Circulating platelet-derived microparticles are increased in severe P . falciparum malaria and serve as a biomarker for neurological involvement [13] , [25] . The murine PbA model has also provided ample evidence for a contribution of CD8+ and CD4+ T cells to the late stages of ECM development [26]–[30] . Both CD8+ T cells , generally considered the terminal effector cells , and CD4+ T cells must accumulate in the cerebral microvasculature for ECM to occur [11] , [27] , [28] , [31]–[35] and may also be responsible for the ECM-associated leukocyte infiltration [36] . While ECM development was thought to involve CD8+ T cell-induced endothelial apoptosis via perforin- and granzyme B-mediated cytotoxicity resulting in BBB disruption [32] , [33] , [37] , we recently showed by IVM that ECM closely correlates with widespread opening of the BBB and that this occurs in the absence of significant endothelial death [21] . The BBB at the level of postcapillary venules encompasses two layers , the vascular endothelium with its basement membrane and the glia limitans with associated basement membranes and astrocyte endfeet , which are separated by the perivascular space [38] . This section of the BBB is functionally distinct from other areas of the BBB , for example that at the capillary level , which consists of a single layer composed of endothelia , gliovascular membrane , and astrocyte endfeet [38] . IVM also revealed that ECM correlates with platelet deposition , leukocyte arrest , and de novo expression of the pattern recognition receptor CD14 on the endothelial surface from postcapillary venules , but not from capillaries or arterioles [21] . Strikingly , inhibition of platelet deposition and leukocyte recruitment by blockage of LFA-1 mediated cellular interactions prevented ECM and disruption of the BBB in PbA-infected mice [21] . Thus , it appears that the ultimate cause of coma and death in ECM is a universal breakdown of the BBB at the level of postcapillary venules [21] . In the PbA-infected CBA/CaJ mouse model , vascular leakage , neurological signs , and death from ECM can be prevented by treatment with the endothelial barrier-stabilizing sphingosine 1 analog FTY720 ( fingolimod ) [21] , [39] , an immunomodulatory FDA-approved drug for oral treatment of relapsing multiple sclerosis ( MS ) [40] that acts as an agonist for sphingosine 1-phosphate ( S1P ) receptors [41] . In experimental autoimmune encephalomyelitis ( EAE ) , FTY720 prevents T cell recruitment to the brain by down-modulating the expression of S1P1 receptors on the T cell surface . This favors the CCR7-mediated retention of naïve and central memory T cells within secondary lymphatic tissues [42] , leading to a reduction in the numbers of naïve and central memory T cells , but not effector memory T cells , in the blood [43] . FTY720 may also prevent stimulation of vascular endothelia or activation of CD8+ effector T cells in the spleen by decreasing CD11c+ DC migration and function and by destabilizing DC/T cell interactions thus preventing the formation of an immunological synapse [44] , [45] . In addition to its involvement in T cell activation and targeting to the brain , FTY720 is also thought to have a directly stabilizing effect on endothelial junctions at the BBB [46]–[49] . However , the exact mechanism by which FTY720 prevents BBB opening remains unclear to date . Here , we show that ECM is associated with the accumulation of numerous leukocytes within postcapillary and larger venules and that the resulting microrheological alterations severely restrict the venous blood flow . Treatment with FTY720 significantly reduced the recruitment of these leukocytes indicating their involvement in the pathogenesis of ECM [21] , [39] . Leukocyte arrest likely increases the intracranial pressure , similarly to P . falciparum iRBC sequestration in pediatric HCM , which is typically associated with a poor clinical outcome [50] .
IVM revealed that the venous blood flow in postcapillary venules from PbA-infected mice with neurological signs ( day 6 ) was strikingly altered . Vascular labeling with Evans blue revealed that postcapillary venules from mice with ECM exhibited a marginal zone devoid of RBCs ( Figure 1A , Video S1 ) . Instead , this zone contained variable numbers of leukocytes that were either rolling along the endothelium , crawling , or firmly attached ( Figure S2A and S2C , Video S2 ) . Minimal projections of time sequences emphasize the boundary between the functional lumen in the center of the postcapillary venules and the RBC-free marginal zone and suggest that the functional lumen available for the blood flow is significantly restricted during ECM ( Figure S2B and S2D ) . This phenomenon was even more pronounced in larger venules . Neither arterioles from mice with ECM ( Figure 1B , Video S3 ) nor postcapillary venules or arterioles from mice with hyperparasitemia ( Figure 1C and 1D , Video S4 ) showed any significant functional vascular restriction , i . e . narrowing of the passageway available for the blood flow , compared to uninfected control mice ( Figure 1E and 1F , Video S5 ) , a finding we attribute to the absence of steric hindrance generated by adherent leukocytes in these vessels . Multiple measurements of the total vascular diameter ( from endothelium to endothelium ) and the functional diameter ( used by the blood flow ) of 50 randomly chosen postcapillary venules from 4 mice with ECM revealed a mean functional diameter of 70 . 5±13 . 7% compared to data from 3 uninfected control mice , corresponding to a functional vascular cross-section of 55 . 8±19 . 1% ( Figure 1G ) . Notably , complete vascular occlusion , whether in postcapillary venules or other microvessels , was not observed during ECM . In 50 randomly chosen postcapillary venules from 3 PyXL-infected mice with hyperparasitemia , the functional postcapillary venule diameter and cross-section was 95 . 5±3 . 4% and 92 . 7±5 . 1% , respectively . No microrheological alterations were found in postcapillary venules or arterioles from PbA-infected mice prior to ECM ( day 5 ) , in PbA-infected mice that failed to develop ECM ( day 9 ) , or in uninfected control mice . As reported previously [21] , PbA-infected mice that did develop ECM despite treatment FTY720 exhibited vascular leakage suggesting that the venous blood flow restriction was similar to untreated PbA-infected mice with ECM . Thus , ECM correlates with a significant functional constraint , but not complete blockage , of the passageway available for the venous blood flow . This is significant , because any restriction in the venous efflux from the brain likely exacerbates edema formation , a hallmark of both ECM and HCM [16] , [17] , [53] . A venous efflux problem would also explain the increased intracranial pressure , which is frequently observed in pediatric CM in Africa [50] . Indeed , MRI imaging has identified increased intracranial pressure as the strongest predictor of death [54] , [55] . Quantitative offline IVM analysis of the various cell densities revealed that leukocytes are recruited to the cortical microvasculature not only in response to ECM , but also hyperparasitemia , albeit at a significantly lower density ( Figure 2 ) . Specifically , more CD8+ T cells , neutrophils , and macrophages were recruited in ECM compared to hyperparasitemia , while the density of all other cell types analyzed did not differ between these two infections ( Figure 3A ) . The cortical microvasculature of uninfected control mice exhibited virtually no arrested leukocytes suggesting that in the absence of an inflammatory stimulus , innate immune cells do not monitor the BBB . To elucidate the composition of specific cellular subtypes involved in pathology we quantified leukocytes by flow cytometry in perfused whole brains . In contrast to IVM , no significant difference in CD8+ T cell recruitment was observed between ECM and HP suggesting that flow cytometry may lack the sensitivity to distinguish important focal variations in cellular composition as arrested leukocytes were not observed in other vessels such as capillaries or arterioles . Overall , significantly more CD45+ leukocytes were found in the brains during ECM compared to hyperparasitemia ( 23729 . 3±7573 . 8 vs . 4483 . 0±2971 . 6; ANOVA: F ( 2 ) = 12 . 42; P<0 . 001 , Tukey's test: T = −4 . 49; P<0 . 05 ) and PbA/FTY720 mice ( 6059 . 7±4070 . 7; Tukey's test: T = −4 . 12; P<0 . 05 ) ( Figure 3C , Table S7 ) . Confirming the IVM data , ECM was associated with the recruitment of significantly higher numbers of Ly6G+ neutrophils than in hyperparasitemia ( 1470 . 7±325 . 5 vs . 518 . 0±317 . 2; ANOVA: F ( 2 ) = 9 . 33; P<0 . 05 ) , while there was no significant difference in the number of Ly6C+ monocytes . The largest increase in cell numbers was observed for F4/80+ macrophages during ECM compared to hyperparasitemia ( 9648 . 0±3432 . 1 vs . 938 . 0±645 . 9; ANOVA: F ( 2 ) = 14 . 35; P<0 . 01 ) . Equivalent results were obtained for CD11b+ macrophages ( ANOVA: F ( 2 ) = 6 . 50; P<0 . 01 ) . Because total brain leukocytes necessarily contain a large proportion of parenchymal macrophages , we distinguished these from blood-derived macrophages by their low level of CD45 expression [36] , [56] . When the CD45lo parenchymal macrophages ( microglia ) were excluded , the number of the remaining mostly intravascular CD45hi F4/80+ macrophages was significantly higher during ECM compared to PyXL-infected mice ( 1750 . 0±285 . 0 vs . 243 . 3±171 . 5; ANOVA: F ( 2 ) = 30 . 19; P<0 . 01 ) ( Figure 3D ) . FTY720 treatment of PbA-infected mice significantly reduced the number of Ly-6G+ neutrophils ( Tukey's Test: T = −3 . 69; P<0 . 05 ) , and both total and CD45hi F4/80+ macrophages ( Tukey's test: T = −4 . 31; P<0 . 05 ) in PbA-infected mice so that no significant difference was found for any of the cell types between PbA/FTY720 mice on day 9 and PyXL-infected mice with hyperparasitemia on day 5 ( Figure 3B-D , Table S7 ) . Equivalent results were obtained for CD11b+ macrophages ( Table S7 ) . Because neither FTY720-treated PbA-infected mice nor PyXL-infected mice with hyperparasitemia exhibit vascular leakage or neurological signs , it appears that FTY720 prevents BBB opening and the associated leukocyte recruitment , although we cannot exclude that FTY720 affects the brain directly . Flow cytometry revealed two distinct leukocyte subsets , namely CD45hi and CD45lo ( Figure 4A ) , both of which were significantly more numerous in the brains of PbA-infected mice ( ANOVA: F ( 2 ) = 10 . 38; P<0 . 05 and ANOVA: F ( 2 ) = 27 . 21; P<0 . 01 , respectively ) compared to PyXL-infected mice ( Tukey's: T = −4 . 45; P<0 . 05 and Tukey's test: T = −7 . 37; P<0 . 01 , respectively ) ( Table S8 ) . FTY720 treatment significantly reduced the number of both CD45hi ( Tukey's test: T = −3 . 08; P<0 . 05 ) and CD45lo ( Tukey's test: T = −3 . 99; P<0 . 05 ) leukocytes . While the number of CD45hi cells after FTY720 treatment was not statistically different from PyXL infection , the number of CD45lo cells , although significantly decreased compared to PbA infection , remained significantly higher compared to PyXL-infected mice with hyperparasitemia ( Tukey's test: T = −3 . 38; P<0 . 05 ) ( Table S8 ) . Significantly more ICAM-1+ ( Table S9 ) and CD69+ leukocytes ( Table S10 ) were present in the CD45hi and the CD45lo leukocyte subsets from PbA-infected mice compared to PyXL-infected mice ( Figure 4B–D ) . FTY720 treatment significantly reduced the number of ICAM-1+ , CD69+ , and GrB+ leukocytes compared to PbA-infected mice with ECM ( Table S9 , S10 , and S11 ) . Further , the CD45hi subset from PbA-infected mouse brains contained consistently higher numbers of ICAM-1+ , CD69+ , and GrB+ CD8+ T cells compared to PbA/FTY720 mice , although this difference was not statistically significant ( Figure S4A , Table S12 ) . Furthermore , the median expression levels of ICAM-1 , CD69 and GrB in these CD45hi CD8+ T cells were similar amongst PbA-infected , PbA/FTY720 , and PyXL-infected mice ( Figure S4B-D ) . Likely , the ECM-associated vasculopathy is caused by the high density of activated leukocytes in the postcapillary venules . P-selectin release from platelet α-granules or endothelial Weibel-Palade bodies promotes the binding of platelets , leukocytes , and plasma proteins to the vascular wall [71]–[73] . Because both platelet marginalization and P-selectin expression have been implicated in the pathogenesis of both HCM and ECM [18] , [20]–[22] , [74]–[78] , we determined the distribution of this adhesion molecule with respect to arrested platelets in the cortical microvasculature . Upon manifestation of neurological signs , PbA-infected CBA/CaJ mice were inoculated with a PE-conjugated mAb against P-selectin ( CD62P ) , eFluor 450-conjugated anti-CD41 to detect platelets and Evans blue to visualize the vascular lumen [21] . IVM revealed small clusters of marginalized platelets that colocalized with patches of P-selectin on cortical postcapillary venule endothelia ( Figure 7 , Video S20 ) . Occasionally , we observed strings of platelets that appeared to be attached to clusters of platelets ( Video S21 ) as has been suggested to occur in HCM based on in vitro experiments [79] . In contrast , PyXL-infected mice with hyperparasitemia showed no evidence for P-selectin expression or platelet arrest ( Figure 7 , Video S22 ) . Unlike postcapillary venules , arterioles were consistently negative for P-selectin or arrested platelets , both during ECM and hyperparasitemia . Thus , ECM , but not hyperparasitemia , is associated with marginalization of small numbers of platelets along postcapillary venule endothelia and P-selectin release , either from platelets or endothelia . However , the highly focal nature of both platelet arrest and P-selectin release contrasts with the uniform endothelial activation as evidenced by CD14 expression , ICAM-1 upregulation , and vascular leakage observed during ECM . Thus , leukocyte arrest is not limited to the P-selectin positive portions of the postcapillary venule endothelia . FTY720 was previously shown to prevent vascular leakage , neurological signs , and death from ECM [21] , [39] . To evaluate whether FTY720 protects the BBB by preserving the integrity of endothelial junctions [48] , [80] , [81] , we determined the expression level of the tight junction ( TJ ) proteins claudin-5 , occludin , and ZO-1 in the cerebral cortex and the cerebellum of 4 PbA-infected mice with ECM ( day 6–8 ) , 3 FTY720-treated PbA-infected mice that did not exhibit any neurological signs ( day 8 or 9 ) , and 3 PyXL-infected mice with hyperparasitemia ( day 5 ) ( Figure S6 and S7 ) . Quantification of the fluorescence emission of specific antibodies on 3–4 immunolabeled cryostat sections per experimental condition yielded no significant reduction in protein expression under the different infection and treatment conditions compared to 3 uninfected control mice ( Table S16 ) . This finding suggests that the TJs remained morphologically intact and supports the hypothesis that the ECM-associated vascular leakage is based on a regulated , potentially reversible , mechanism of BBB opening [21] , [82] . Thus , comparison of two Plasmodium infection models revealed: 1 ) The venous blood flow impairment during ECM is caused by the arrest of significantly higher numbers of CD8+ T cells , neutrophils , and in particular macrophages in cortical postcapillary venules compared to hyperparasitemia . While a small number of CD8+ T cells and macrophages extravasated into the perivascular space , most of the recruited leukocytes remained intravascular . 2 ) FTY720 treatment of PbA-infected mice reduced , but did not completely prevent leukocyte accumulation in postcapillary venules , which is consistent with the finding that low numbers of arrested leukocytes are present in PyXL-infected mice with hyperparasitemia without causing vascular leakage or neurological signs . 3 ) ECM closely correlates with the recruitment of large numbers of ICAM-1 expressing F4/80+ macrophages to the brain . As FTY720 treatment did not reduce the ICAM-1 expression level , the high density of these macrophages in postcapillary venules likely enhances the ECM-associated vascular pathology . 4 ) Leukocyte recruitment coincides with the onset of neurological signs , but follows BBB opening , as vascular leakage can be observed 1 day prior to symptomatic ECM [21] .
In this study , we identify a novel key determinant of ECM pathogenesis , namely that leukocyte arrest along the wall of postcapillary venules causes microrheological alterations that severely impair the venous blood flow . Based on our findings , we hypothesize that infection with PbA opens the BBB , which leads to the recruitment of numerous activated CD8+ T cells , ICAM-1+ macrophages , and neutrophils ( Figure 8 ) . The resulting steric hindrance of the blood flow in postcapillary and larger venules impairs , but does not block , the venous efflux from the brain , which exacerbates the vasogenic edema and causes death as a consequence of intracranial hypertension . Under physiological conditions , the luminal surface of vascular endothelia is covered with a glycocalix , a 0 . 5 to >1 µm layer of membrane-bound proteoglycans and glycoproteins that repels RBCs and is critically involved in inflammatory responses , blood coagulation , and blood flow regulation [83]–[86] . IVM visualizes this glycocalix as a thin red layer , covering arteriolar endothelia from infected and uninfected mice and postcapillary venule endothelia from uninfected control mice . During ECM , the thickness of the RBC-free layer in postcapillary and larger venules was drastically increased . Because the glycocalix typically degrades under inflammatory conditions , leading to exposure of adhesion molecules , leukocyte adhesion , and impairment of endothelial barrier function [85] , [86] , the restriction in the venous blood flow during ECM is likely not caused by components of the glycocalix , but by increasing numbers of arrested leukocytes that prevent RBC from approaching the endothelium . Although the functional cross-section of postcapillary venules was occasionally reduced by more than 80% , complete vascular obstruction was not observed . These findings argue in favor of a combined vascular sequestration and immuno-pathological etiology of ECM [87]–[89] . The reduction in the venous blood flow must be expected to have major consequences for the physiology of the brain . First , the overall hypoperfusion of the brain , enhanced by inadequate contact between RBCs and the endothelium , likely contributes to the drastically reduced O2 delivery to the cerebral parenchyma observed in ECM-susceptible C57BL/6 mice [90] . In addition , by increasing the wall shear stress , leukocyte adhesion is expected to reduce the blood volume flow in postcapillary venules dramatically [91] . Finally , a reduction of the venous efflux from the skull , caused by a generalized narrowing of the lumen of venous microvessels , necessarily increases the intracranial pressure . The finding that brains from mice with ECM , but not hyperparasitemia , are swollen and spongy and bulge out of the skull , if the Dura mater is accidentally damaged during craniotomy clearly documents the dramatically increased intracranial pressure during the agonal phase of the disease . The reduced venous efflux from the brain may exacerbate vascular leakage , brain edema , and hemorrhages - cerebral alterations that are also associated with HCM [92] . Brain swelling and edema is extremely common in adult HCM on CT scan [93] , [94] . Increased intracranial pressure has long been associated with poor prognosis and neurological sequelae in severe pediatric HCM [50] , [95]–[98] . In fact , recent longitudinal MRI observations in Malawian children have identified intracranial hypertension as the single most important MRI finding associated with HCM development and the most reliable predictor of death [54] , [99] . PbA-infected mice also exhibit arteriolar vasospasms during the final stage of ECM [100] and it has been suggested that the reduction in the cerebral blood flow observed by MRI [101] is due to increased production of vasoconstrictive factors or inhibition of vasodilating mediators [102] . Subsequent work [103] , [104] revealed that endothelin-1 ( ET-1 ) , a potent vasoactive peptide with inflammatory and platelet-activating properties [105]–[108] , is upregulated during both ECM and HCM [109]–[112] . Indeed , the arteriolar vasoconstrictive effect of ET-1 could be responsible for ECM induction in the PbA-infected C57BL/6 mouse model [108] , because injection of exogenous ET-1 induces neurological signs in PbNK65-infected mice , which normally do not develop ECM [113] , and because blockage of the ET-1 receptor A prevents ECM development in PbA-infected mice [110] . However , ET-1 has a plasma half-life of well under one minute in rodents [114] so that vasoconstriction alone cannot explain the increased intracranial pressure observed during ECM . Because ET-1 also stimulates endothelial activation with upregulation of adhesion molecules , promotes leukocyte adhesion , and increases vascular permeability [105] , [106] , there is a possibility that ET-1 induces ECM by restricting the venous blood flow . Similarly , administration of nitric oxide ( NO ) , a key messenger involved in regulation of platelet adhesion and inflammatory and immune responses [115] , decreased both leukocyte accumulation and vascular resistance in larger venules of PbA-infected mice [78] , [116]–[121] . We conclude that the ultimate cause of death from ECM is a combination of arteriolar vasoconstriction and severe reduction in the blood efflux from the brain due to leukocyte adhesion in the venous microvasculature . Compared to PyXL-infected mice with hyperparasitemia , PbA infection triggered the recruitment of significantly more CD8+ T cells to postcapillary venules at the time of , but not prior to , ECM development . Together with the finding that CD8+ T cells were absent in mice that survived the critical time for ECM development , these data suggest that neurological signs and T cell recruitment are correlated and occur rapidly . CD8+ T cells from C57BL/6 mice immunized with PyXNL can confer protection against lethal PyXL infection [122] , whereas CD8+ T cell accumulation in the brain of PbA-infected C57BL/6 mice was abolished and the mice were completely protected from ECM when co-infected with P . yoelii [123] , [124] . However , similar numbers of CD8+ T cells accumulated in the brains of PbA-infected C57BL/6 mice with ECM and PbNK65-infected mice without neurological signs [34] , further supporting the notion that ECM-eliciting parasites such as PbA induce the recruitment of a qualitatively different CD8+ T cell population to the brain . FTY720 treatment decreased the number of CD8+ T cells to levels similar to those found in PyXL-infected mice . Despite the presence of the remaining CD8+ T cells , neither FTY720-treated PbA-infected mice nor PyXL-infected mice developed neurological signs . A small percentage of CD8+ T cells entered the perivascular space during ECM , but not hyperparasitemia . This could be explained by upregulation of the leukocyte common antigen CD45 , because CD45 expression is typically enhanced in response to stress signals , leading to increased leukocyte motility [125] and brain infiltration , for example after seizure [126] . ECM coincided with larger numbers of CD8+ T cells expressing CD69 , one of the earliest lymphocyte activation markers [127] , [128] , and FTY720 treatment reduced the number of CD69+ CD8+ T cells to levels similar to those found during hyperparasitemia . Previous work supports a correlation between ECM and CD69+ CD8+ T cells: 1 ) Recruitment and activation of CD8+ T cells and CD69 expression were reduced in ECM-resistant mice [129] . 2 ) Peripheral CD8+ T cells were predominantly CD69+ during ECM and expressed the phenotype of memory T cells [130] . 3 ) The ECM-associated upregulation of CD69 was reversed and disease was prevented by interference with the angiotensin I pathway [131] . 4 ) Ghanaian P . falciparum infected pediatric patients with clinical HCM or severe anemia showed similar T cell activation profiles with a significantly increased frequency of CD69+ cells compared to asymptomatic children [132] . The median expression levels per cell of CD69 and GrB did not differ between the experimental groups suggesting that ECM pathogenesis correlates with a high density of CD69+ and GrB+ CD45hi CD8+ T cells in postcapillary venules . FTY720 treatment of PbA-infected mice reduced the number of CD4+ T cells to levels similar to those observed for PyXL-infected mice . Together , these data suggest that FTY720 treatment prevents ECM by inhibiting the trafficking of activated CD8+ and CD4+ T cells to the brain . FTY720 treatment of PbA-infected mice also reduced the number of macrophages and neutrophils to levels similar to those found in the brains of PyXL-infected mice , supporting the notion that these leukocytes exacerbate edema formation during ECM [26] , [27] . Of particular importance , FTY720 treatment prevented the recruitment of large numbers of ICAM-1+ macrophages to the brains of PbA-infected mice . This finding sheds light on an unexpected new role of ICAM-1 in the pathogenesis of ECM . While FTY720 may preserve the integrity of the BBB primarily by preventing leukocyte recruitment to the brain , activated platelets release phosphorylated FTY720 [133] , [134] , which acts as a full agonist on the endothelial S1P receptor S1PR4; therefore , FTY720 may prevent the ECM-associated vascular leakage by strengthening the endothelial actin cytoskeleton [135] , [136] . Further , FTY720 may regulate endothelial barrier function by directly modulating endothelial junction tightness , because the increased vascular leakage observed in mice deficient in plasma S1P can be reversed by restoring plasma S1P levels [47] , [48] , [80] , [81] . This finding may explain why FTY720 administration must be started prior to the onset of vascular leakage [21] , [39] and why attempts to rescue mice with symptomatic ECM , when leukocyte recruitment was already in progress , were unsuccessful ( A . Movila and U . Frevert , unpublished observations ) . Further , FTY720 can cross the BBB and may thus be able to directly modulate parenchymal cells by interacting with S1P receptors in the CNS . The resulting feedback from the CNS on the activation status of the BBB may in turn alter the interaction with the immune cells . Future testing of FTY720 or related compounds in the ECM model is expected to reveal more detail on the exact mechanism of BBB opening and vascular inflammation in ECM . The ECM model may also improve our understanding of the pathogenesis of HCM . Ugandan , Malawian , and Central Indian children with HCM exhibit decreased S1P plasma levels compared to those with uncomplicated malaria and a low angiotensin-1 to angiotensin-2 plasma ratio discriminates HCM and severe non-cerebral from uncomplicated malaria and also predicts mortality from HCM [39] , [137]–[140] . As these reports strongly suggest the involvement of the S1P pathway HCM , screening for novel immunomodulatory drugs and exploration of their endothelial barrier-promoting effects is warranted . We observed significantly more neutrophils in postcapillary venules and whole brain during ECM compared to hyperparasitemia . Considering that the role of neutrophils in the pathogenesis of CM is understudied to date , this finding is of particular interest . FTY720 treatment of PbA-infected mice reduced the number of neutrophils significantly so that levels similar to those found during hyperparasitemia were reached , which supports the previously suggested involvement of neutrophils in the ECM-associated vasculopathy and edema formation [26] , [141] . The number of arrested monocytes was increased similarly in PbA- and PyXL-infected mice compared to uninfected control mice suggesting that monocyte recruitment correlates with Plasmodium infection in general , not ECM in particular . This finding is in agreement with earlier reports showing that monocytes are not involved in iRBC accumulation in the brain at the time of ECM [26] , [27] . Intravenous injection of fluorescent anti-CD14 revealed that monocytes were generally confined to the vascular lumen , although we occasionally detected labeled monocytes in the Virchow-Robin space . Similarly , intravenous injection of the macrophage marker anti-CD11b resulted in labeling of PVM . Because part of the PVM population derives from blood monocytes [53] , these fluorescent monocytes may have been labeled before extravasating to replenish the pool of PVM in the perivascular space . Flow cytometric measurement of leukocyte recruitment to the brain essentially corroborated our IVM findings suggesting that the cortical microvasculature reflects the ECM-associated pathological events in the entire brain . A few conceptual differences between IVM and flow cytometry are noteworthy . First , IVM confirmed the notion that the healthy murine brain has an extremely low level of immune surveillance with the almost complete absence of T cells , neutrophils , monocytes , and B cells [21] , [142] . Expression of adhesion molecules is low on the healthy human postcapillary venule endothelium , and upregulation of P-selectin , E-selectin , ICAM-1 , and VCAM-1 contributes to the pathogenesis of HCM [53] , [58] , [143] . Second , IVM provides information on the location of recruited leukocytes and their dynamics with respect to the complex and heterogeneous microvascular network of the brain . Third , intravenous labeling generally visualizes only intravascular leukocytes including those that extravasate between marker injection and imaging . However , the BBB at the level of postcapillary venules is naturally leaky , i . e . it allows passage of macromolecules including immunoglobulins into the Virchow-Robin space [38] . During ECM , this barrier becomes even more permeable , which makes distinction of recently extravasated from resident perivascular cells difficult . While CD8+ T cells do not patrol the cerebral microvasculature of uninfected mice and are therefore unlikely to extravasate into the perivascular space of naïve mice , the fluorescent CD11b+ macrophages we found in the perivascular space of mice with ECM could either have been labeled intravascularly and then extravasated or they could have been labeled after entering the perivascular space due to diffusion of the cell-specific markers across the BBB . Thus , while the majority of leukocytes remain confined to the vascular lumen during ECM , a small number of CD8+ T cells and possibly CD11b+ macrophages extravasate into the perivascular space . Fourth , flow cytometric measurement of the expression level of CD45 allowed distinction between blood-derived and parenchymal macrophages [56] . Using this approach , we found that ECM is associated not only with significantly higher numbers of ICAM-1+ CD45hi ( = blood-derived ) macrophages , but also ICAM-1+ CD45lo ( = parenchymal ) macrophages compared to mice with hyperparasitemia and FTY720-treated PbA-infected mice . Together with the ECM-associated activation of microglia [144] and the enhanced expression of CCR5 on non-hematopoietic cells [145] , ICAM-1 upregulation on both blood-derived and parenchymal macrophages supports the notion of a mixed vasogenic and cytotoxic nature of the edema in ECM [17] , [146] . Finally , IVM was critical to visualize endothelial activation markers within defined parts of the microvascular tree . Thus , IVM and flow cytometry provided highly complementary results . BBB opening in the young CBA/CaJ mice used here begins at least one day prior to the onset of neurological signs , while leukocyte recruitment to the brain occurs only afterwards [21] . In 129 , B6 mice , neither antibody-mediated neutrophil depletion nor clodronate-mediated macrophage depletion , done 1 or 2 days prior to ECM development , respectively , prevented neurological signs [26] , [145] . While this was interpreted as neutrophils and macrophages not being involved in ECM development , an alternative explanation is that the BBB was already leaky at this late time point and the resulting vasogenic edema had already caused microglial activation . Clodronate liposomes do not cross the BBB [147]–[149] so that the microglia was likely unaffected and the resulting cytotoxic edema not prevented in these studies . Further , depletion of macrophages and granulocytes by administration of AP20187 to MAFIA mice on days 5 , 6 , and 7 after infection resulted in a >80% reduction of these cells in the blood [27] , [150] . Again , as AP20187 does not cross the BBB [151] , ECM development likely manifested because the activated microglia was not eliminated . Importantly , CD8+ T cells alone failed to induce any of the typical signs of ECM including convulsions and death without the direct neurotoxic effect of intravenously administered folic acid [26] , [145] . Thus , it appears that both hematogenic and parenchymal macrophages play a role of in the pathogenesis of ECM . A more recent study in C57BL/6 mice emphasizes the crucial role of monocytes/macrophages in lymphocyte recruitment to the brain . Clodronate treatment 2 days prior to manifestation of neurological signs reduced the recruitment of CD8+ T cells , CD4+ T cells , and NK cells to the brain 2 . 8-fold , 1 . 8-fold , and 4 . 6-fold , respectively , and failed to alter the course of disease development [152] . Clodronate treatment 2 days prior to infection with PbA , on the other hand , prevented ECM development [152] . Thus , the exact mechanism by which blood-derived macrophages contribute to disease development needs further and more specific investigation . Taken together , these data suggest the following scenario: PbA infection induces BBB opening , which results in vasogenic edema and microglial activation . The ensuing cytotoxic edema then leads to endothelial activation with recruitment of leukocytes that in concert restrict the venous blood flow , which further weakens the already impaired BBB . Eventually , these events culminate in the typical irreversible damage associated with fatal malarial encephalopathy . However , the differential progression of the vasogenic versus the cytotoxic edema and their relative contribution to brainstem compression and death clearly require further experimental attention . The critical role of ICAM-1 in the pathogenesis of severe P . falciparum malaria and the binding of P . falciparum iRBC to the endothelium is generally accepted [59] , [61] , [153]–[159] . However , comparison of the endothelial ICAM-1 expression levels in the brains of P . falciparum versus P . vivax infected individuals is necessary to determine whether or not HCM correlates with upregulation of ICAM-1 in postcapillary venules . ICAM-1 upregulation has also been implicated in the pathogenesis of ECM [63] , [77] , [78] , [160]–[162] . While the cellular origin of ICAM-1 was not determined , treatment of PbA-infected mice with NO reduced ICAM-1 upregulation in the brain along with leukocyte sequestration and vascular leakage [78] . We show that upregulation of endothelial ICAM-1 correlates with Plasmodium infection in general , not with ECM in particular . Because ICAM-1 upregulation was not observed in mice deficient in RAG-1 or IFN-γ on day 6 after infection with PbA [163] , the Th1 type cytokines IFN-γ and lymphotoxin , in synergy with TNF from macrophages [163] , [164] , may induce ICAM-1 expression throughout the brain microvasculature of ECM-susceptible mice , both during ECM and hyperparasitemia . Together with the finding that FTY720 treatment of PbA-infected mice does not prevent ICAM-1 upregulation in postcapillary venules , these data argue against a role of this endothelial adhesion molecule in the pathogenesis of ECM . CD14 appeared on the surface of postcapillary venule endothelia selectively at the time of neurological signs ( day 6–8 ) , which is in agreement with the well-known involvement of CD14 in endothelial activation , neuroinflammation , and leukocyte recruitment to the cerebral microvasculature [165]–[168] and the finding that CD14-deficient mice are protected against ECM [169] . Further , because CD14 plays an important role in the non-phlogistic clearance of apoptotic cells [170] , its expression on postcapillary venule endothelia may reflect the well-documented increase in apoptosis during ECM [171]–[173] . Interestingly , malaria-associated microparticles carry phosphatidylserine on their surface [174] . Because the plasma levels of endothelial , platelet , and erythrocytic microparticles increase at the onset of the ECM-associated neurological signs [175] , CD14 may be involved in the recruitment of microparticles to postcapillary venule endothelia . Platelet-derived microparticles are known to promote macrophage differentiation [176] and may contribute to the observed upregulation of ICAM-1 on the macrophage surface . We show that ECM occurs without apparent physical degradation of endothelial TJs , measurable loss of claudin-5 , occludin , and ZO-1 and in the absence of endothelial death [21] . Together , these findings argue against widespread and irreversible endothelial injury , for example due to cytotoxic T cell-mediated apoptosis or long-lasting vascular occlusion [177] , as a major pathogenetic mechanism leading to malarial coma and death [82] , [88] , [178] , [179] . Instead , and in agreement with the observation that ECM-susceptible mice can be rescued by anti-LFA-1 treatment even minutes before imminent death [180] , [181] , the preserved TJ integrity suggests that BBB opening during ECM is under the control of a regulated mechanism [21] , [82] , [182] . In support of this notion , fast-acting anti-malarial drugs can prevent ECM one day before the expected onset of neurological signs , although CD8+ T cells still accumulate in the brain [32] , [34] . Platelets play a crucial role in the early stages of the pathogenetic cascade of ECM [9] , [183] , [184] . Blockage of platelet activation or shedding of platelet-derived microparticles confers resistance to ECM [13] , [22] , [160] , [184] . At the time of ECM , platelets arrest focally and in small numbers in the cortical microvasculature [21] where they closely associate with P-selectin positive areas on postcapillary venule endothelia , confirming the reported involvement of both platelets [18] , [20]–[22] , [74]–[76] and P-selectin [77] , [78] in the pathogenesis of HCM and ECM . Although platelets produce many factors including VEGF , ROS , and thrombin that are able to directly impair endothelial barrier function [185] , the highly focal nature of platelet arrest [20] , [21] is difficult to reconcile with the homogenous vascular leakage throughout the entire venous microvasculature at the time of ECM [21] . In agreement with their crucial role early in the pathogenesis of ECM , platelets are more likely to accomplish their vascular permeability-augmenting effect indirectly , namely by 1 ) facilitating leukocyte arrest in postcapillary venules [22] , [57] , [186] , [187] , 2 ) boosting ROS formation in leukocytes [185] , and 3 ) enhancing cytokine secretion and cytotoxic capacity of effector T cells [188] . Thus , platelets adhering to small foci of endothelial P-selectin may serve as early nucleation points for the gradually spreading , cytokine-induced vasculopathy observed in postcapillary venules from ECM-susceptible mice . In conclusion , we find that steric hindrance , mediated by large numbers of arrested CD8+ T cells , macrophages , and neutrophils in postcapillary venules , causes a severe restriction in the venous blood flow . Based on this observation , we hypothesize that completely different pathogenetic mechanisms - cytoadherence of P . falciparum iRBC in HCM and sequestration of leukocytes in ECM - can result in the same pathophysiological outcome: a severe reduction in the blood efflux from the brain . If the resulting increase in intracranial pressure intensifies the cerebral edema to the point of brainstem herniation , then compression of the respiration centers in pons and medulla could cause death by respiratory arrest . Intracranial hypertension is a known risk factor for poor outcome in pediatric HCM [189] , [190] , but how this leads to neural injury is unknown [191] , [192] . As discussed in a recent review [193] , intracranial hypertension eliminates the need to explain any selective recognition mechanism Plasmodium might use to target multiple sensitive sites in the brain [194] . Intracranial hypertension leading to general swelling and hypoperfusion of the brain can also explain all of the neurological sequelae in HCM survivors [194] . Reports associating loss of smell , deafness , and blindness with both HCM and ECM support the notion that sequestration of iRBC as well as leukocytes can cause intracranial hypertension [192] , [195]–[200] . Further , leukocyte sequestration is likely also involved in intracranial hypertension during P . falciparum HCM , as artesunate treatment was more efficacious in Asian adults compared to African children with more mononuclear cell accumulation [157] , [201] , [202] and also failed to rescue HCM patients with a low parasite biomass in the brain [203] . Thus , HCM and ECM induce very similar neurological symptoms and sequelae . Thus , despite fundamental differences in parasite biology , the non-cytoadherent rodent parasite PbA could be used as a model to better understand how rheological alterations might lead to the annual death from HCM of over half a million people [204] . Perhaps more importantly , venous efflux disturbances due to leukocyte sequestration could also explain the cerebral complications that are increasingly reported for severe infections with P . vivax , another knobless ( essentially non-cytoadherent ) Plasmodium species [205]–[214] . The marked proinflammatory responses and reversible microvascular dysfunction associated with P . vivax infections [215] , [216] , together with the shared propensity for reticulocyte invasion [217]–[220] , may render the PbA-infected mouse model suitable for study of the pathogenesis of severe P . vivax malaria . Hence , this model promises to shed light on the ultimate cause ( s ) of death from cerebral malaria .
This study was conducted 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 , NYU School of Medicine ( Protocol number 101201-01 ) . All surgery was performed under ketamine-xylazine-acepromazine anesthesia , and all efforts were made to minimize suffering . P . berghei and P . yoelii parasites were maintained by passage through female Anopheles stephensi mosquitoes [221] . The green fluorescent P . berghei ANKA strain ( PbA-GFP ) was a kind gift from Dr . Andy Waters , University of Glasgow , UK [222] . The lethal P . yoelii strain 17XL , originally derived from the non-lethal 17X strain [223] , was kindly provided by Dr . James Burns ( Drexel University College of Medicine ) [224] , [225] . WT PyXL were transfected to express RedStar , an improved version of the red fluorescent protein drFP583/DsRed/RFP [226] , under the control of the elongation factor 1α promoter using a novel replacement strategy [21] , [227] , [228] and termed PyXL-RFP [21] . Both PbA-GFP and PyXL-RFP emit fluorescence throughout the entire life cycle . Mice were CBA/CaJ ( Jackson Laboratory , Bar Harbor , ME ) . Animals were maintained and used in accordance with recommendations in the guide for the Care and Use of Laboratory Animals . CBA/CaJ mice were infected at the age of 3 weeks ( body weight of 12–15 g ) by intraperitoneal injection of 0 . 5–10×106 iRBC as described [21] . In our hands , 3 week-old CBA/CaJ mice responded to PbA infection with neurological signs and died from ECM comparable to adult mice used by others [3] , [6] , [18] , [22] , [23] , [76] , [141] , [169] , [229]–[244] . The parasitemia was monitored daily using Giemsa stained blood smears and mice were sacrificed upon development of ECM or hyperparasitemia . PbA-infected mice were considered ECM positive when two or more parameters clearly indicated behavioral alteration including body position , spontaneous activity , startle response , tremor , gait , touch escape , and righting reflex [245] . Quantitative assessment of ECM-associated neurological signs was performed using the Rapid Murine Coma and Behavior Scale ( RMCBS ) with values of 3–7 defined here as severe , 8–16 as mild , and 17–20 as no ECM [246] . PyXL-infected mice exhibited a hunched position , pale skin color , and an increased respiration rate when parasitemia levels reached levels around 80% , but failed to present typical neurological manifestations . For IVM , groups of 15 PbA-infected CBA/CaJ mice received one daily oral dose of 0 . 3 mg/kg FTY720 starting one day before infection or no treatment as described [21] , [39] . At the onset of neurological signs , mice were injected with Evans blue and examined by IVM . Mice surviving the critical period of ECM development were inoculated with Evans blue and PE-conjugated species anti-species CD8a and imaged on day 9 . Groups of 5 PyXL-infected or uninfected mice were inoculated with the same markers and imaged on day 5 for comparison . For flow cytometry , PbA-infected mice , FTY720-treated PbA-infected mice , and PyXL-infected mice were analyzed on day 6–8 , day 8 , and day 5 , respectively . Mice were anesthetized by intraperitoneal injection of a cocktail of 50 mg/kg ketamine ( Ketaset , Fort Dodge Animal Health , Fort Dodge , IO ) , 10 mg/kg xylazine ( Rompun , Bayer , Shawnee Mission , KS ) , and 1 . 7 mg/kg acepromazine ( Boehringer Ingelheim Vetmedica , St . Joseph , MO ) ( KXA mix ) and surgically prepared for intravital imaging of the brain as described [21] , [247] , [248] . CBA/CaJ mice were infected with PbA and subjected to brain IVM 1 ) on day 5 prior to the appearance of neurological signs , 2 ) on day 6–8 upon ECM development , or 3 ) on day 9 after the window of ECM development had passed . PyXL-infected mice were imaged upon the parasitemia exceeding 50% on day 5 . Uninfected mice were used as controls . Prior to imaging , mice were inoculated with Evans blue and matching combinations of fluorescent markers . Despite severe illness of the animals , optimization of anesthesia , craniotomy , and injection of fluorescent markers allowed us to obtain good recordings from approximately 70% of the mice with ECM and 50% of the mice with hyperparasitemia . iRBC were identified by fluorescent protein expression in the parasites or reflection of hemozoin [21] . The vascular lumen was visualized by intravenous injection of 100 µl of a 1% solution of Evans blue . Vascular endothelia were labeled intravenously with Alexa 488 or eFluor 450-conjugated rat anti-mouse PECAM-1 ( CD31; clone MEC13 . 3 , BioLegend , San Diego , CA ) and phycoerythrin ( PE ) - or Alexa 647-conjugated rat anti-mouse CD14 ( clone Sa2-8 , eBioscience ) . CD8+ T cells , CD4+ T cells , monocytes , macrophages , neutrophils , or platelets were labeled by intravenous injection of 3–5 µg of the following fluorochrome-conjugated monoclonal antibodies using appropriate color-matching combinations: PE-conjugated rat anti-mouse CD8a ( clone 53-6 . 7; eBioscience , San Diego , CA ) , eFluor 450 or PE-conjugated rat anti-mouse CD4+ ( clone GK 1 . 5 , eBioscience ) , Pacific blue-conjugated rat anti-mouse CD11b ( clone M1/70 , BioLegend ) , eFluor 450- , PE- or Alexa 647-conjugated rat anti-mouse GR-1 ( Ly-6G/6c; clone RB6-8C5 , eBioscience and BioLegend ) , and eFluor 450-conjugated rat anti-mouse CD41 ( clone MWReg30 , eBioscience ) , respectively . ICAM-1 expression was visualized with intravenously inoculated PE-conjugated rat anti-mouse CD54 ( clone YN 1/1 . 7 . 4 , Biolegend ) . P-selectin was detected with PE-conjugated rat anti-human/mouse CD62p ( KO2 . 3 , eBioscience ) . Multiple time sequences and 3D stacks were recorded for quantification of the number of arrested leukocytes in the vascular lumen . Depending on the experimental conditions , 20–45 postcapillary venules and arterioles were analyzed per mouse . Arrested leukocytes were defined in each vessel segment as cells that did not detach from the endothelial lining within the observation period . CD8+ T cells , CD4+ T cells , neutrophils , monocytes , and macrophages were quantified by counting the number of cells per square millimeter of vessel surface [249] . The relative density of the various leukocyte populations was determined in multiple fields of view per experimental condition and expressed as the mean ± SEM of arrested cells as well as the percentage of the total cell number . Velocities were measured with Imaris Track as described [250] . To quantify endothelial ICAM-1 expression , confocal 3D stacks of postcapillary venules or similarly sized arterioles were acquired from 2 mice each infected with PbA , PyXL , or no parasites . The fluorescence signal intensity across 10 ( PbA ) or 12 ( PyXL and control ) vessel volumes was collected from 3D data and quantified in ImageJ . To quantify leukocyte ICAM-1 expression , the relative fluorescence emission from individual leukocytes was measured from mice infected with PbA or PyXL . As for endothelial ICAM-1 expression , confocal 3D stacks were collected and projections were created in AutoDeblur . A total of 6 stacks from 2 mice per experimental condition were analyzed . Leukocytes were prepared from the brain ( cerebrum and cerebellum ) using established procedures [251]–[253] . Briefly , mice were perfused intracardially with Mg2+ and Ca2+-free PBS to dislodge nonadherent leukocytes . Next , the brain was removed and gently minced through a mesh strainer ( mesh size: 100 µm; Fisher Scientific ) using a syringe plunger . The homogenate was suspended in 10 ml HBSS containing 0 . 05% collagenase D ( Roche Diagnostics , Indianapolis , IN ) , 0 . 1 µg/ml of the trypsin inhibitor TLCK ( Sigma ) , 10 µg/ml DNase I ( Sigma ) , and 10 mM Hepes buffer , pH 7 . 4 . The tissue slurry was gently rocked at RT for 60 min and then allowed to settle at 1 g for 30 min . The supernatant was collected and 5 ml of the suspension was layered onto 10 ml of a density separation medium containing 7 . 5 ml of RPMI medium containing 10% FBS , 10 mM HEPES , and 2 . 5 ml Ficoll Paque ( GE HealthCare ) in a 50 ml conical centrifuge tube and centrifuged at 400 g for 30 min . The overlying media and tissue debris were removed , the entire gradient medium was diluted ten-fold with HBSS and centrifuged at 300 g for 10 min . Isolated leukocytes were washed twice with Mg2+ and Ca2+-free PBS before phenotyping . A total of 10 PbA-infected mice with ECM , 6 PbA-infected/FTY720-treated , and 6 PyXL-infected mice with hyperparasitemia were subjected to flow cytometric analysis . Uninfected control mice were not included in the analysis , because the cerebral microvasculature of these animals does not exhibit arrested leukocytes . The following antibodies were used for leukocyte phenotyping: Total leukocytes were detected with PE-Cy7-conjugated rat anti-mouse CD45 ( clone 30-F11; eBioscience , San Diego , CA ) , T cells with APC-Cy7-conjugated Armenian hamster anti-mouse CD3 ( clone 145-2C11; BD Biosciences , San Jose , CA ) , CD8+ T cells with eFluor 450 or Alexa 700-conjugated rat anti-mouse CD8a ( clone 53-6 . 7; eBioscience or Biolegend , San Diego , CA ) , CD4+ T cells with PE- Texas Red-conjugated rat anti-mouse CD4+ ( clone RM4-5; Life Technologies , Carlsbad , CA , neutrophils with FITC-conjugated rat anti-mouse Ly6G ( clone 1A8; Biolegend ) , monocytes with Alexa 700-conjugated rat anti-mouse Ly6C ( clone HK 1 . 4; Biolegend ) , and macrophages with anti-mouse CD11b ( clone M1/70; eBioscience ) or PE- or eFluor 450-conjugated anti-mouse F4/80 ( clone BM8; eBioscience ) . CCR5 ( CD195 ) was detected with PE-conjugated rat anti-mouse CD195 ( clone HM-CCR5; Biolegend ) , CD69 with Pacific Blue-conjugated rat anti-mouse CD69 ( clone H1 . 2F3; Biolegend ) , ICAM-1 with APC-conjugated rat anti-mouse CD54 ( clone YN1/1 . 7 . 4; eBioscience ) , and granzyme B with FITC-conjugated anti-mouse granzyme B ( clone GB11; Biolegend ) . Data were acquired with a 5-laser , 17-color LSR II Analytic Flow Cytometer ( BD Biosciences , San Jose , CA ) and analyzed with FlowJo software ( Treestar , Ashland , OR ) . Time series showing the blood flow in postcapillary venules or similarly sized arterioles were acquired from mice infected with PbA , PyXL , or no parasites and converted to minimal projections to visualize the portion of the vascular lumen used for blood flow . The vascular lumen was visualized by IV inoculation of Evans blue [21] . IVM movies were converted to minimal projections to visualize the perfused center of the vessel . Multiple measurements were taken for each vessel to determine the entire vascular diameter ( distance between endothelia ) versus the perfused part of the vessel ( dark center ) ( Figures 1 and S2 ) . Vascular restriction is expressed as percent reduction in vessel diameter or cross-section . The cortical microvasculature was imaged with an inverted Leica TCS SP2 AOBS confocal system as described [21] . Time series and 3D data sets were acquired with Leica Confocal Software [248] , [250] , [254] . Imaris 7 . 4 ( Bitplane , Saint Paul , MN ) , Image-Pro Plus ( Media Cybernetics , Bethesda , MD ) , AutoDeBlur ( Media Cybernetics , Bethesda , MD ) , and NIH ImageJ were used for further image analysis , deconvolution , and 3D reconstruction . Blood was removed by perfusion with PBS via the left ventricle [251] . Brains were snap-frozen for preparation of cryostat sections . To determine the expression level of tight junction proteins , sections were fixed in 95% ice-cold ethanol for 30 minutes and then permeabilized in acetone for 1 min at RT . After blocking in 50% goat serum for 45 min , sections were labeled with rabbit polyclonal antibody against claudin 5 ( 34-1600 ) , rabbit polyclonal anti-occludin ( 40-4700 ) , mouse monoclonal anti-ZO1 ( 33-9100 ) , all from Invitrogen . Rat monoclonal antibody anti-CD31 ( 553370 ) was from BD Biosciences . Sections were then incubated for 2 h at RT with secondary Alexa antibodies ( Life Technologies ) and mounted in Vectashield with DAPI ( H1200 , Vector Labs , Burlingame , CA ) . Depending on the experimental condition , 9–16 images from various regions of the cerebrum and cerebellum were analyzed . Data were acquired at the same magnification with Zeiss Axiovison software using an Axio Imager-D2 Zeiss fluorescent microscope and imported into Image J for quantification of vascular leakage and tight junction protein expression . For tight junction protein expression , the fluorescence threshold was set with the “triangle” setting of Image J to limit the measurement of the brightest signal only , corresponding to the tight junctions . Ratios were then calculated by comparing the experimental groups with the control group using Excel . Significance was determined by t-test . Statistical analysis was performed using Minitab® 17 ( Minitab , State College , PA ) . Data sets were tested for normality and equal variances and when necessary data was Log10 transformed to obtain a normal distribution . Data were analyzed either by t-test , one-way ANOVA , or General Linear Model as appropriate . Where transformation was not successful , the non-parametric Mann-Whitney U test was used . | Malaria remains one of the most serious health problems globally , but our understanding of the biology of the Plasmodium parasite and the pathogenesis of severe disease is still limited . Human cerebral malaria ( HCM ) , a severe neurological complication characterized by rapid progression from headache to convulsions and unrousable coma , causes the death of hundreds of thousands of children in Africa annually . To better understand the pathogenesis of cerebral malaria , we imaged immune cells in brain microvessels of mice with experimental cerebral malaria ( ECM ) versus mice with malarial hyperparasitemia , which lack neurological impairment . Death from ECM closely correlated with plasma leakage , platelet marginalization , and the recruitment of significantly more leukocytes to postcapillary venules compared to hyperparasitemia . Leukocyte arrest in postcapillary venules caused a severe restriction in the venous blood flow and the immunomodulatory drug FTY720 prevents this recruitment and death from ECM . We propose a model for ECM in which leukocyte arrest , analogous to the sequestration of P . falciparum infected red blood cells in HCM , severely restricts the venous blood flow , which exacerbates edema and swelling of the brain at the agonal comatose stage of the infection , leading to intracranial hypertension and death . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"apicomplexa",
"parasite",
"groups",
"medicine",
"and",
"health",
"sciences",
"plasmodium",
"biology",
"and",
"life",
"sciences",
"cerebral",
"malaria",
"malaria",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitology"
] | 2014 | Experimental Cerebral Malaria Pathogenesis—Hemodynamics at the Blood Brain Barrier |
HLA-B*27 exerts protective effects in hepatitis C virus ( HCV ) and human immunodeficiency virus ( HIV ) infections . While the immunological and virological features of HLA-B*27-mediated protection are not fully understood , there is growing evidence that the presentation of specific immunodominant HLA-B*27-restricted CD8+ T-cell epitopes contributes to this phenomenon in both infections . Indeed , protection can be linked to single immunodominant CD8+ T-cell epitopes and functional constraints on escape mutations within these epitopes . To better define the immunological mechanisms underlying HLA-B*27-mediated protection in HCV infection , we analyzed the functional avidity , functional profile , antiviral efficacy and naïve precursor frequency of CD8+ T cells targeting the immunodominant HLA-B*27-restricted HCV-specific epitope as well as its antigen processing and presentation . For comparison , HLA-A*02-restricted HCV-specific epitopes were analyzed . The HLA-B*27-restricted CD8+ T-cell epitope was not superior to epitopes restricted by HLA-A*02 when considering the functional avidity , functional profile , antiviral efficacy or naïve precursor frequency . However , the peptide region containing the HLA-B*27-restricted epitope was degraded extremely fast by both the constitutive proteasome and the immunoproteasome . This efficient proteasomal processing that could be blocked by proteasome inhibitors was highly dependent on the hydrophobic regions flanking the epitope and led to rapid and abundant presentation of the epitope on the cell surface of antigen presenting cells . Our data suggest that rapid antigen processing may be a key immunological feature of this protective and immunodominant HLA-B*27-restricted HCV-specific epitope .
The human leukocyte antigen ( HLA ) B*27 is associated with a high rate of spontaneous viral clearance in hepatitis C virus ( HCV ) infection [1] , [2] and with slow disease progression in human immunodeficiency virus ( HIV ) infection [3] , [4] . In both infections , the protective role has been linked to single immunodominant CD8+ T-cell epitopes [5]–[9] . Virological and immunological mechanisms contribute to the HLA-B*27-mediated protection . For example , extensive virological studies have demonstrated that viral escape from CD8+ T-cell responses that target the protective immunodominant HLA-B*27-restricted epitopes in both HIV and HCV infection is difficult to achieve and requires the accumulation of several mutations . In HIV infection , three mutations within and outside the immunodominant HLA-B*27-restricted epitope ( KK10 ) are required for viral escape: the first mutation has an immunmodulatory effect , a second mutation compensates for viral fitness costs , and a third mutation abrogates HLA binding [10]–[13] . In HCV infection , viral escape mutations are not tolerated at the HLA-B*27 binding anchors of the epitope due to a major impact on viral replicative fitness . Mutations at T-cell-receptor contact residues can occur in this otherwise highly conserved region; however , several of these mutations are required for full escape due to broad T-cell cross-recognition of viral variants [14] . Thus , in both infections , related but distinct mechanisms constrain virological escape and contribute to protection . In contrast , the immunological mechanisms that contribute to HLA-B*27-mediated protection are less well understood . It is possible that certain characteristics of HLA-B*27-restricted CD8+ T cells may contribute to the protective effect . Indeed , functional avidity , defined as the sensitivity of CD8+ T cells to antigenic stimulation , has been proposed to correlate with the outcome of viral infection [15]–[19] . For example , one study found that the functional avidity of HLA-B*27-restricted HIV-specific CD8+ T-cell responses directed against the protective immunodominant KK10 epitope was higher in comparison to responses directed against epitopes restricted by other HLA-alleles [15] . Others reported that the functional avidity of KK10-specific CD8+ T-cell responses was comparable to that of responses targeting subdominant HLA-B*27-restricted epitopes derived from HIV [18] . It has also been suggested that immunodominant HLA-B*27-restricted HIV-specific CD8+ T-cell responses display more potent antiviral efficacy compared to responses targeting subdominant HLA-B*27-restricted epitopes [18] . Another factor that has been suggested to contribute to the superior control of HIV replication is the polyfunctionality of the immunodominant KK10-specific CD8+ T cells [15] . Polyfunctionality is characterized by the simultaneous production of multiple effector-molecules such as CD107a , IFN-γ , IL-2 , MIP-1β and TNF-α . However , since HLA-B*27-mediated protection is also closely linked to the immunodominance of HLA-B*27-restricted virus-specific CD8+ T-cell responses , it is tempting to speculate that factors determining immunodominance may help explain this protective effect . Immunodominance is expected to be determined early after infection when naïve antigen-specific CD8+ T cells encounter their antigen and the first responses start to emerge . Factors that contribute to CD8+ T-cell immunodominance include antigen processing and presentation , abundance of peptide-major histocompatibility complex class I ( pMHCI ) molecules on antigen-presenting cells ( APCs ) and the number of naïve T cells that express complementary T-cell receptors [20] . Studies primarily performed in mouse models have shown that the immunodominance hierarchy and the magnitude of T-cell responses are shaped by the number of antigen-specific naïve CD8+ T cells [21]–[23] . Consistent with these findings , we have shown that the immunodominance hierarchy of HLA-A*02-restricted HCV-specific CD8+ T-cell responses observed in chronically HCV infected subjects is related to the frequency of naïve CD8+ precursors [24] . However , there is also growing evidence that immunodominance is largely influenced by differences in antigen processing [25]–[28] . One study has demonstrated that antigen processing strongly influences CD8+ T-cell response hierarchies in HIV infection as the amount of epitope produced correlated with CD8+ T-cell response magnitude and frequency [29] . Moreover , proteasomal cleavage of HIV Gag-derived polypeptides by the immunoproteasome efficiently generated precursors of the HIV-specific HLA-B*27-restricted protective immunodominant KK10 epitope as well as minute amounts of the optimal epitope itself [29] , [30] . The ‘optimal’ epitope is defined as the epitope form that results in the greatest in vitro stimulation of epitope-specific CD8+ T-cell responses . In this comprehensive study , we set out to elucidate the immunological factors that mediate the protection afforded by HLA-B*27 in HCV infection . Compared to HLA-A*02-restricted HCV-specific CD8+ T-cell responses , we found that the protective effect of the immunodominant HLA-B*27-restricted HCV-specific epitope is not clearly linked to differences in the intrinsic properties of cognate CD8+ T-cell populations , but rather to more rapid and efficient antigen processing and presentation . These results suggest that both immunodominance and protection in this case are associated with the kinetics and efficacy of epitope generation .
In a first set of experiments , we assessed the functional avidity of HCV-specific CD8+ T-cell populations from chronically HCV infected subjects . For comparative purposes , CD8+ T-cell populations specific for three frequently recognized HLA-A*02-restricted epitopes were evaluated in addition to those specific for the immunodominant HLA-B*27-restricted NS5B2841 epitope ( Table 1 ) . All four epitopes bind comparably and with high affinity to their restricting HLA class I molecule [31] . In HLA-A*02+ tested subjects only one or a maximum of two epitope-specific CD8+ T cell responses were detectable and thus analyzed . It is important to note , that in all functional assays performed in this study , short term expanded HCV-specific CD8+ T-cell lines ( 14 days ) were used . Circulating HCV-specific CD8+ T cells in chronically infected subjects are only present at a very low frequency ( usually <0 . 1% of total CD8+ T cells ) ; this does not allow the performance of functional analyses . After two weeks of peptide-specific expansion , cytotoxic T lymphocyte ( CTL ) lines were stimulated with HLA-B*27+ or HLA-A*02+ subject-derived EBV-immortalized B-cell lines pulsed with serial peptide concentrations ( Figure 1A ) ; the response EC50 was quantified as the exogenous peptide concentration required to yield half-maximal frequencies of cells producing intracellular IFN-γ , as shown in Figure 1B . Importantly , HLA-B*27-restricted CD8+ T cells displayed a significantly lower mean functional avidity ( indicated by a significantly higher mean EC50 value ) compared to HLA-A*02-restricted CD8+ T cells from the same subject group ( Figure 1C ) . Since the presence of chronic infection may affect the properties of virus-specific CD8+ T-cell populations , we also analyzed the functional avidity of CD8+ T cells obtained from subjects with resolved HCV infection . Interestingly , HLA-B*27-restricted CD8+ T cells from subjects with resolved HCV infection displayed a higher mean functional avidity compared to cells obtained from chronically infected HLA-B*27+ subjects although this difference was not statistically significant ( Figure 1C ) . However , this did not exceed the mean functional avidity observed for HLA-A*02-restricted epitopes targeted in chronic or resolved infection . Thus , superior functional avidity is not a characteristic of HLA-B*27-restricted CD8+ T cells specific for the protective immunodominant NS5B2841 epitope when compared to HLA-A*02-restricted CD8+ T-cell populations . In order to compare the functional profile of HLA-B*27- and HLA-A*02-restricted CD8+ T cells from chronically HCV-infected subjects we performed intracellular multi-cytokine staining . Upon peptide-specific stimulation , HLA-B*27- and HLA-A*02-restricted CTL lines produced the same array of cytokines with even similar quantities ( Figure 2A ) . Particularly , CD107a was mobilized and IFN-γ , MIP-1β and TNF-α were produced by CD8+ T cells restricted by both HLA-alleles . However , production of IL-2 by all virus-specific CD8+ T cells was limited ( Figure 2A ) and neither HLA-B*27- nor HLA-A*02-restricted CD8+ T cells produced IL-4 , IL-10 , IL-17A or IL-22 ( data not shown ) . The direct number of functions ( polyfunctional profile ) of individual CD8+ T cells is shown in Figure 2B and 2C . Although CD8+ T cells restricted by both alleles produced similar amounts of the same cytokines , individual HLA-B*27-restricted CD8+ T cells produced more cytokines simultaneously compared to HLA-A*02-restricted CD8+ T cells . Importantly , they showed a higher proportion of CD107a+ , IFN-γ+ and MIP-1β+ cells . These results suggest that HLA-B*27- and HLA-A*02-restricted CD8+ T cells derived from chronically infected subjects have similar functional qualities with comparable quantities even though their polyfunctional profile at a cell based level may slightly differ . Next , we analyzed whether HLA-B*27-restricted NS5B2841 epitope-specific CD8+ T cells exert superior antiviral efficacy compared to CD8+ T cells of other specificities . In HIV infection , it has been suggested that a protective immunodominant HLA-B*27-restricted CD8+ T-cell response is characterized by superior antiviral activity [18] . To address this possibility , we used human hepatoma HuH7 cells harboring a JFH1-based selectable subgenomic luciferase replicon to measure the inhibition of HCV replication by CD8+ T cells . These replicon cells were transduced with lentiviral vectors expressing HLA-A*02 or HLA-B*27 with a selectable marker conferring blasticidin resistance controlled by the constitutive cellular EF1a promoter , as described previously [32] . The resulting cell lines showed strong and homogeneous expression of HLA-A*02 or HLA-B*27 , respectively , by flow cytometric analysis ( data not shown ) . The inhibitory effect on HCV replication was measured by determining luciferase activity , which is fully dependent on HCV RNA replication and correlates precisely with intracellular levels of viral RNA antigens [32] . Since the subgenomic replicon is based on the genotype 2a isolate JFH1 and its sequence is not cross-recognized by the respective CD8+ T cells specific for genotype 1a peptides , pulsing with respective peptide is required ( genotype 1a sequence , see Table 1 ) . HuH7 cells expressing HLA-A*02 were pulsed with increasing concentrations of specific HLA-A*02-restricted peptides and HuH7 cells expressing HLA-B*27 were pulsed with increasing concentrations of the dominant HLA-B*27 peptide . The target cells were then cocultured with respective peptide-specific CTL lines at an effector-to-target ( E∶T ) ratio of 1∶1 . Importantly , as shown in Figure 3 , virus-specific CD8+ T cells blocked HCV replication in a peptide dose-dependent manner that correlated precisely with the amount of IFN-γ secreted by the virus-specific T cells . Of note , HLA-B*27-restricted CD8+ T cells specific for the NS5B2841 epitope derived from chronically infected subjects did not mediate superior antiviral efficacy . HLA-A*02-restricted CD8+ T cells started to inhibit HCV replication at a lower peptide concentration ( 10−8 ) compared to HLA-B*27-restricted CD8+ T cells ( Figure 3A and B ) . We additionally analyzed the antiviral efficacy of CD8+ T cells derived from subjects with resolved HCV infection ( Figure 3C and D ) . HLA-B*27-restricted CD8+ T cells from subjects with resolved HCV infection displayed a higher antiviral efficacy compared to cells obtained from chronically infected HLA-B*27+ subjects . However , they were not much superior to HLA-A*02-restricted CD8+ T cells obtained from subjects with chronic or resolved infection . These findings show that the antiviral efficacy is correlated with the functional avidity of CD8+ T cells and suggest that the protective effect of the immunodominant HLA-B*27-restricted epitope is not explained by the antiviral efficacy of epitope-specific CD8+ T cells . Next , we asked whether the strong immunodominance of the NS5B2841 epitope might be dictated by a high frequency of naïve CD8+ T-cell precursors . To address this issue , we analyzed naïve precursor frequencies of CD8+ T cells specific for the NS5B2841 epitope and compared it to the well described immunodominant HLA-A*02-restricted NS31406 epitope . Of note , we recently showed that the immunodominance of this latter epitope is linked to a relatively high frequency of epitope-specific naïve precursors [24] . By using a previously described combination of tetramer staining , magnetic-bead enrichment and multiparametric flow cytometry [33] , we were able to detect naïve HCV-specific CD8+ T cells in all four HLA-B*27 and all eight HLA-A*02 healthy donors analyzed ( Figure 4 ) . Although the cells were not detectable before the enrichment step , they could be clearly identified after enrichment , as shown for one representative subject in Figure 4A . The cells displayed a naïve phenotype characterized by high expression of CD45RA , CD27 and CCR7 , and low expression of CD11a ( data not shown ) . Of note , the naïve precursor frequency of HLA-B*27-restricted CD8+ T cells specific for NS5B2841 was not significantly higher compared to the frequency of the immunodominant HLA-A*02-restricted naïve precursors ( Figure 4B , data of HLA-A*02-restricted naïve precursors were previously published [24] ) . These data suggest that the strong immunodominance of the HLA-B*27-restricted NS5B2841 epitope is not due to an intrinsically high naïve precursor frequency , at least across HLA restriction elements . Recently , it has been shown that antigen processing influences HIV-specific CD8+ T-cell response hierarchies [29] . In case of an HLA-B*27-restricted HIV Gag-derived epitope , immunodominance correlated with abundant proteasomal production of a range of short and long , naturally processed peptide forms containing the optimal epitope [29] . We therefore analyzed whether proteasomal cleavage of the dominant NS5B2841 epitope has an influence on the immunodominant response to this epitope . Specifically , we performed a proteasomal digest of a 25-amino acid peptide containing the HLA-B*27-restricted dominant NS5B2841 epitope , as well as two peptides containing the HLA-A*02-restricted epitopes NS31073 and NS5B2594 ( Figure 5A ) , using constitutive proteasomes and immunoproteasomes in parallel over a time course of six hours . Subsequently , the resulting fragments were analyzed by mass spectrometry . Our initial proteasomal digestion experiments resulted in production of the optimal HLA-A*02-restricted NS31073 and NS5B2594 ( A2-NS31073 and A2-NS5B2594 ) epitopes , but surprisingly failed to yield any of the HLA-B*27-restricted NS5B2841 ( B27-NS5B2841 ) epitope or even B27-NS5B2841 epitope-containing fragments ( Figure 5B ) . Indeed , the 25-mer B27-NS5B2841 polypeptide was completely degraded and only peptides of four to nine amino acids in length could be detected ( Figure 5C ) . We therefore performed additional digestion experiments with this polypeptide substrate at a four-fold dilution of both proteasomal forms . By using this more sensitive approach , we were able to detect small amounts of the optimal B27-NS5B2841 epitope after four to six hours of immunoproteasomal digestion . Importantly , we also observed six additional longer peptides containing the B27-NS5B2841 epitope , each of which was present at concentrations exceeding that of the optimal B27-NS5B2841 epitope by up to 300-fold ( Figure 5D; the optimal B27-NS5B2841 epitope as well as all B27-NS5B2841 epitope-containing fragments are highlighted in black ) . Because some of these fragments ended at the C-terminal end of the analyzed 25-mer peptide , and production thus was a maximum estimate , we performed additional proteasomal digestions of an overlapping , but more C-terminal 25-mer peptide encompassing B27-NS5B2841 . These proteasomal digestions demonstrated cleavage C-terminal of the first analyzed fragment ( Figure S1 ) . Although epitope production was less than indicated by the maximum estimates of our first experiment , the overall production of the B27-NS5B2841 epitope-containing fragments still comprised between 8–32% of all digestion products ( data not shown ) . To investigate whether the rapid proteasomal processing was due to the B27-NS5B2841 sequence itself , or to the very hydrophobic regions flanking this epitope , we generated an artificial chimeric 25-mer peptide containing the B27-NS5B2841 epitope imbedded in the regions flanking the A2-NS5B2594 epitope ( termed B27-in-A2 ) . We found that this B27-in-A2 peptide degraded in a manner more similar to the original A2-in-A2 25-mer peptide than the original B27-in-B27 peptide ( Figure 6 ) . Indeed , the chimeric B27-in-A2 peptide degraded slightly more slowly than the A2-in-A2 peptide following immunoproteasomal digestion and although we doubled the amount of proteasome used in this experiment , we could not increase the rate of degradation to that of the original B27-in-B27 peptide . Consequently , the hydrophobicity of the B27-NS5B2841 epitope flanking region is critical to the rapid processing of this epitope . Our data furthermore suggests that the combination of epitope and flanking region affect the overall peptide degradation rate . In order to analyze whether during degradation of the chimeric B27-in-A2 peptide the optimal epitope is generated , we compared the amount of optimal 9-mer B27-NS5B2841 epitope generated following digestion of the artificial chimeric B27-in-A2 peptide with that made following processing of the natural B27-in-B27 peptide ( Figure S2 ) . Our data show that more of the optimal B27-NS5B2841 epitope was released following both constitutive and immunoproteasomal digestion when the epitope was in the A2-NS5B2594 flanking context , although the epitope was produced more slowly . It is important to note , that we compared only the amounts of optimal 9-mer epitope . Longer versions of the B27-NS52841 epitope are not produced following processing of the artificial chimeric B27-in-A2 construct while these are made in great amounts when the 25-mer peptide with the natural B27-in-B27 amino acid sequence is processed . As it has been shown previously that HLA-B*27 is able to bind and present peptides that are N- or C-terminally extended fragments of an optimal epitope [29] , [34] , we next examined the cross-recognition of the seven B27-NS5B2841 epitope-containing peptide forms using non-specifically expanded CD8+ selected T cells derived from two chronically HCV infected HLA-B*27+ subjects by intracellular cytokine staining for IFN-γ . Importantly , these CD8+ T cells responded to almost all naturally processed B27-NS5B2841 epitope-containing peptides ( Figure 7A and B ) . In one subject , some of the long peptide variants induced even slightly stronger IFN-γ production compared to the optimal B27-NS5B2841 epitope ( data not shown ) . In these experiments , we used non-specifically expanded CD8+ T cells rather than peptide-stimulated T-cell lines because they best represent the range of CD8+ T-cell populations found in subjects in vivo . To confirm specificity , we also analyzed the capability of the NS5B2841 epitope-containing peptide forms to stimulate the expansion of epitope-specific CD8+ T cells derived from chronically infected HLA-B*27+ subjects; all tested peptide forms were able to stimulate epitope-specific CD8+ T-cell expansions in culture ( Figure 7C ) . The rapid proteasomal cleavage of the HLA-B*27-restricted NS5B2841 epitope prompted us to investigate whether this leads to faster epitope presentation on the cell surface of APCs . Since it is hypothesized that the timing of antigen abundance at the cell surface of antigen presenting cells during initial priming of naive HCV-specific CD8+ T may influence the immunodominace hierarchy of responding CD8+ T cells [25] , [35] and since it is assumed that APCs rather than hepatocytes initially prime naïve HCV-specific CD8+ T cells ( reviewed by [36]–[38] ) , we decided to use subject-derived APCs for our assays . Due to the lack of approved models of expanded subject-derived DCs we decided to use EBV-immortalized B-cell lines that best fit to this model of antigen processing and presentation . HLA-B*27+ and HLA-A*02+ EBV-immortalized B-cell lines were infected with vaccinia virus constructs encoding HCV proteins containing the relevant peptides . Importantly , the sequence of the vaccinia virus construct is identical to the sequence of the polypeptides used for our proteasomal digestion experiments ( genotype 1a ) . Subsequently , after infection the cells were cultured for different incubation times ( 0–24 h ) to allow endogenous antigen processing , then added to peptide-specific CTL lines restricted by HLA-B*27 or HLA-A*02 , respectively ( Figure 8A ) . Induction of IFN-γ production as a read-out for antigen presentation was measured by intracellular cytokine staining after five hours of cocultivation . As shown for representative subjects in Figure 8B and C and for all subjects tested in Figure 8D , antigen-specific CD8+ T cell IFN-γ production became detectable within two to four hours of vaccinia virus infection , suggesting that endogenously processed epitopes had already reached the surface of the APC in this time period . Importantly , the induction of IFN-γ in HLA-B*27-restricted CD8+ T cells was higher at two and four hours after infection compared to HLA-A*02-restricted CD8+ T cells , even despite the lower avidity of the HLA-B*27-restricted CD8+ T-cell responses ( Figure 8D ) . When corrected for functional avidity in individual subjects ( Figure S3 ) , a significantly higher level of the HLA-B*27-restricted NS5B2841 peptide was presented at the cell surface compared to HLA-A*02-restricted epitopes at these time points ( Figure 8E ) . Finally , we analyzed the influence of proteasome inhibitors on the endogenous antigen processing and their effect on the timing of the induction of antigen-specific IFN-γ production of HLA-B*27 and HLA-A*02-restricted CTL lines . By using the proteasome inhibitor epoxomicin we were able to delay the induction of epitope-specific IFN-γ production and considerably decrease the number of responding CD8+ T cells ( Figure 8F ) . We also performed control experiments with the proteasome inhibitor lactacystin and saw 30–70% reductions in the number of responding CD8+ T cells ( data not shown ) . These data support the assumption that proteasomal digestion more likely than alternative cytosolic proteases , such as the tricorn , is responsible for the production of HCV-specific epitopes . In sum , these results demonstrate that , following endogenous processing via the proteasome , the HLA-B*27-restricted NS5B2841 epitope more rapidly reaches a higher peptide concentration on the cell surface compared to HLA-A*02-restricted epitopes , and that this translates to kinetically enhanced induction of antigen-specific IFN-γ production .
In this study , we set out to determine the immunological mechanisms that contribute to the protective effect associated with the immunodominant HLA-B*27-restricted NS5B2841 epitope in HCV infection . Importantly , we found that protection cannot be clearly linked to CD8+ T-cell responses characteristics , such as functional avidity or antiviral efficacy . Indeed , it was the first finding of our study that in chronic infection protective immunodominant HLA-B*27-restricted CD8+ T-cell responses had lower functional avidity than HLA-A*02-restricted CD8+ T-cell responses . These results contrast somewhat with previous studies performed in HIV infection , which have observed that CD8+ T-cell responses restricted by HLA-B alleles have a higher functional avidity than those restricted by HLA-A alleles . Most notably , HLA-B*27-restricted CD8+ T-cell responses specific for the immunodominant HIV Gag-derived epitope KK10 , which is associated with slow disease progression , displayed the highest functional avidity when tested using the optimal epitope [15] , [16] . In contrast , however , CD8+ T-cell responses towards longer natural processed peptide forms of this epitope have a much lower functional avidity than those against the optimal epitope [29] . Thus , in the infected host CD8+ T-cell cross-recognition of optimal and natural peptide forms may result in biased in vitro estimations of functional avidity and magnitude of epitope-specific CD8+ T-cell responses when only using the optimal epitope . In HCV infection , Yerly et al . observed higher levels of functional avidity for CD8+ T-cell responses in subjects who cleared the virus compared to CD8+ T-cell responses derived from subjects with chronic HCV infection [19] . Although we also found a higher level of functional avidity for HLA-B*27-restricted CD8+ T-cell responses derived from subjects with resolved versus chronic HCV infection , this functional avidity was still lower compared to HLA-A*02-restricted CD8+ T-cell responses . Collectively , our results suggest that the overall protective effect of HLA-B*27 in HCV infection cannot be linked to superior functional avidity even when we used the optimal HCV-epitope in our experiments . This conclusion is in keeping with a recent study by Harari et al . , who did not find a correlation between functional avidity and protective CD8+ T-cell responses specific for several viruses [17] . In addition to functional avidity , we studied other aspects of the CD8+ T-cell response to the protective immunodominant HLA-B*27-restricted NS5B2841 epitope . Notably , we found that the antigen-specific naïve CD8+ T-cell precursor frequency for this epitope was not substantially different compared to an immunodominant HLA-A*02-restricted epitope . This is particularly interesting in the light of previous studies suggesting that the number of naïve precursors plays a key role in the generation of immunodominance hierarchies during viral infections in mouse models and humans [21] , [23] , [24] , [39] . However , detailed comparison with subdominant HCV-derived HLA-B*27-restricted epitopes would be required to interpret this observation fully . Furthermore , we found no evidence that CD8+ T-cell responses to the protective immunodominant HLA-B*27-restricted NS5B2841 epitope mediated enhanced antiviral efficacy . This observation is somewhat perplexing when searching solely for an immunological explanation for the protective effect of the HLA-B*27-restricted NS5B2841 epitope , but gains credence in the context of combined virological considerations . In addition , it should be noted that this parameter was assessed in the context of peptide-pulsed cell lines due to system constraints . In the context of naturally presented epitopes , it remains feasible that rapid and efficient antigen processing could translate into enhanced antiviral efficacy in vivo . We finally analyzed the functional profile of HCV-specific CD8+ T cells and found that HLA-B*27- and HLA-A*02-restricted CD8+ T cells have the same combination of functions with similar amounts of produced effector-molecules . However , it is important to note that HLA-B*27-restricted CD8+ T cells have a higher capacity to simultaneously produce multiple effector-molecules compared to HLA-A*02-restricted CD8+ T cells . This is in line with studies performed with HIV-specific CD8+ T cells which have shown that CD8+ T cells specific for the immunodominant HIV KK10-epitope display a superior polyfunctional profile compared to CD8+ T cells restricted by other HLA-alleles [15] and that polyfunctionality is generally connected with a superior control of HIV infection [40] . However , differences in polyfunctionality were much lower in our study compared to these studies performed in HIV infection . The most important finding of our study is that the protective effect of the immunodominant HLA-B*27-restricted NS5B2841 epitope can be linked to extraordinarily rapid processing by both proteasomal forms and fast presentation of the epitope at the cell surface . These important data were obtained by different but complementary , experimental approaches . First , by using biochemical assays with purified constitutive proteasomes and immunoproteasomes , we demonstrated that the region containing the HLA-B*27-restricted NS5B284 epitope is rapidly processed resulting in fast generation of epitope precursors . We furthermore show that this rapid processing mainly is due to the very hydrophobic regions flanking the NS5B284 epitope , which is in line with the finding of Lucciari-Hartz et al . who demonstrated that processing of hydrophobic protein regions of HIV was more efficient than that of hydrophilic regions [41] . In this context , it is important to note that proteasomal digestion generated the optimal HLA-B*27-restricted NS5B2841 epitope only in small amounts and mainly produced longer precursors of this epitope . However , these naturally processed longer NS5B2841-containing peptide fragments could be cross-recognized by CD8+ T cells derived from chronically HCV infected subjects . These results indicate that HLA-B*27 may be able to present these extended peptide forms [34] and that optimal epitopes are not necessarily the epitope peptide forms produced most frequently in the infected host [29] . Interestingly , the important role of antigen processing and presentation of protective immunodominant viral epitopes has also been demonstrated recently for the protective immunodominant HLA-B*27-restricted HIV Gag-derived KK10 epitope . Specifically , the optimal KK10 epitope was generated only in minimal amounts while epitope precursors of this epitope were produced efficiently by proteasomal cleavage [29] . Therefore , the predominance of proteasomal production of long peptides comprising the NS5B2841 epitope , rather than the optimal NS5B2841 epitope itself , is similar to the situation observed previously for the KK10 epitope . Likewise , the patterns of CD8+ T-cell cross-recognition and response magnitudes specific for the optimal NS5B2841 epitope and the longer NS5B2841-containing peptides in our experiments are similar to those found in subjects with chronic HIV infection who respond to the KK10 epitope [29] . However , the HLA-B*27-restricted NS5B2841 epitope-containing region is processed significantly faster than the HIV Gag-derived KK10 epitope . Importantly , by using functional T-cell assays , we could show that the rapid processing of the protective immunodominant HLA-B*27-restricted NS5B2841 epitope resulted in early presentation at the surface of APCs . Collectively , these results support the hypothesis that early antigen processing kinetics , rather than absolute epitope quantities , help to elicit the protective immunodominant HLA-B*27-restricted NS5B2841 response . This has not previously been observed in humans , but is in agreement with analyses of CD8+ T-cell response hierarchies in normal and immunoproteasome-deficient mice infected with Listeria monocytogenes . Indeed , without immunoproteasomal processing , the presentation of an otherwise immunodominant epitope was delayed and failed to evoke a CD8+ T-cell response [25] . Thus , the magnitude and kinetics of antigen-specific CD8+ T-cell responses appear to be determined during the first 24 hours after infection . Furthermore , T-cell response hierarchies were defined before the peak of the inflammatory response and prior to substantial bacterial replication [42] . In addition , competition for the same APC between T cells of different specificities has been shown to occur within the first 5 hours of immunization and to affect the number of T cells responding to a specific antigen in mice [43] . These results may in part be due to the highly regulated capture , processing and presentation of antigens by dendritic cells , which may only have a limited window-of-opportunity to present antigen and engage in long , stable interactions with naïve T cells [35] . Thus , epitopes that are processed and presented faster than others may have a relatively greater chance of evoking early and abundant T-cell responses in infected subjects similar to what has been suggested for infected or immunized mice [25] , [35] . This notion is supported by a study showing that the magnitude of HIV-specific CD8+ T-cell responses restricted by HLA-A alleles was dramatically lower in the presence of the protective alleles HLA-B*27 and HLA-B*57 , but not in the presence of other HLA-B alleles [44] . In concert with these studies , our results support the biological relevance of rapid processing and presentation in early viral infection , which may in turn lead to the efficient induction of immunodominant and possibly protective CD8+ T-cell responses . Importantly , the analysis of the proteasomal degradation of an artificial chimeric peptide containing of the optimal B27-NS5B2841 epitope surrounded by the A2-NS5B2594 flanking regions showed that the optimal epitope can be produced in higher amounts in this context compared to that of the natural B27-NS5B2841 epitope-containing peptide . These results suggest that for the design of an HCV vaccine construct the timing but also the amount of epitope processed by the proteasome can be manipulated through modifications of the epitope itself as well as of the epitope flanking regions . As discussed above epitope amount may not be as important as the timing of epitope presentation for the priming of virus-specific CD8+ T cells in acute infection; however it may play a more important role for the priming of CD8+ T cells using a vaccine . Taken together , our findings suggest that immunological factors such as rapid antigen processing and presentation contribute to immunodominance hierarchies and combine with virological factors such as functional constraints on viral escape to generate protective CD8+ T-cell responses in human viral infections such as HIV and HCV . Our results also suggest that HCV immunogens could be modified and optimized in vitro to increase the rate of proteasomal processing and thus the likelihood of evoking abundant , or perhaps even immunodominant , CD8+ T-cell responses towards any epitope . This possibility has clear implications for the design of an HCV vaccine .
Written informed consent was obtained in all cases and the study was conducted in agreement with the 1975 Declaration of Helsinki , federal guidelines and local ethics committee regulations . The ethics committee of the Albert-Ludwigs-Universität , Freiburg approved the study . Twelve subjects with chronic HCV infection ( six HLA-A*02+ and six HLA-B*27+ ) who presented to the University Hospital of Freiburg were included in the study . In addition , six subjects with resolved HCV infection ( two HLA-A*02+ , one double-positive HLA-A*02+/HLA-B*27+ and three HLA-B*27+ ) and eleven healthy individuals were included . Peripheral blood mononuclear cells ( PBMCs ) were isolated from EDTA anticoagulated blood samples using lymphocyte separation medium density gradients ( PAA Laboratories GmbH ) . Procedural details are similar to those described by Alanio et al . [33] . In brief , PBMCs ( 1–2×108 ) were incubated for 15 minutes at 4°C with FcR blocking reagent ( Miltenyi Biotec ) and stained for 30 minutes with pMHCI tetramer conjugated to allophycocyanin ( APC ) , with the pMHCI component at 20 nM final concentration . A small aliquot of the labeled cells was removed for staining ( pre-enriched fraction ) and the remaining cells were incubated for 20 minutes at 4°C with anti-APC microbeads ( Miltenyi Biotec ) . Again , a small aliquot was removed for counting the pre-enriched fraction and the remaining cells were passed over a magnetic-activated cell separation column ( Miltenyi Biotec ) . After removal of the column from the magnet , the bound cell population ( enriched fraction ) as well as the flow-through ( depleted fraction ) were collected and stained . The frequency of the naïve epitope-specific T-cell population was determined using a calculation similar to that of Alanio et al . [33] , specifically the absolute number of phenotypically naïve ( CD45RAhi , CD27hi , CCR7hi , CD11alow ) tetramer+ CD8+ T cells/the absolute number of CD8+ T cells was calculated . Cell populations ( pre-enriched , enriched and depleted fractions ) were labeled with a combination of anti-CD45RA-phycoerythrin ( PE ) , anti-CD27-APC-eFlour780 ( eBioscience ) , anti-CD8-AmCyan , anti-CCR7-PE-Cy7 , anti-CD3-Pacific-Blue and anti-CD11a-fluorescein isothiocyanate ( FITC ) monoclonal antibodies ( mAbs; BD Biosciences ) . Viaprobe ( 7-AAD; BD Biosciences ) was used for the exclusion of dead cells . The cells were stained in PBS supplemented with 5% fetal calf serum ( Pan-Biotech ) for 20 minutes and washed twice before and after addition of the staining reagents . All samples were acquired using a FACS Canto II flow cytometer ( BD Biosciences ) and analyzed with FlowJo software ( TreeStar Inc . ) . Peptides corresponding to immunodominant HCV-epitopes restricted by HLA-A*02 ( CINGVCWTV , KLVALGINAV and ALYDVVSKL ) and HLA-B*27 ( ARMILMTHF ) , as well as long peptide forms of ARMILMTHF , were synthesized with a free amino and carboxy terminus by standard Fmoc chemistry ( Genaxxon Bioscience ) . All peptides were dissolved and diluted according to previously reported protocols [45] . Tetrameric pMHCI complexes were generated as described previously [46] . Procedural details were performed as described previously [9] . In brief , CD8+ T cells were enriched from PBMCs using magnetic beads coupled to anti-CD8 mAbs ( Dynabeads , Dynal ) and a particle magnetic concentrator . The enriched CD8+ fractions were cultured in 2 mL complete medium ( RPMI 1640 ( Gibco ) containing 10% fetal calf serum , 1% streptomycin/penicillin ( Invitrogen ) and 1 . 5% HEPES buffer , 1 mol/L ( Biochrom ) ) supplemented with 100 U/mL recombinant human interleukin-2 ( rIL-2; Hoffmann-La Roche ) , 0 . 04 µg/mL anti-human CD3 mAb ( Immunotech ) and irradiated autologous PBMCs . Twice per week , 1 mL medium supplemented with 200 U/mL IL-2 was exchanged . On day 14 , the expanded CD8+ cells were used for intracellular IFN-γ staining . Experiments were performed as described previously [9] . In brief , PBMCs ( 4×106 ) were resuspended in 1 mL complete medium and stimulated with peptide at 10 µg/mL final concentration in the presence of 0 . 5 µg/mL anti-CD28 mAb ( BD PharMingen ) . On days 3 and 10 , 1 mL complete medium supplemented with rIL-2 at 20 U/mL final concentration was added . On day 7 , the cultures were restimulated with the corresponding peptide ( 10 µg/mL ) and 1×106 irradiated autologous PBMCs . On day 14 , the cells were used for intracellular IFN-γ staining . The recombinant vaccinia virus construct vHCV 827 , which encodes all relevant HCV peptides according to the sequence of genotype 1a , together with a vaccinia virus encoding the T7 RNA polymerase ( vTF7 ) ( both generously provided by Charles Rice , Rockefeller University , New York [47] ) , were used to induce transient expression of endogenously processed HCV peptides in HLA-matched EBV-immortalized B-cell lines ( B-LCLs ) . B-LCLs were infected at a multiplicity of infection ( MOI ) of 50 for 1 hour with vHCV 827 and vTF7 , or with vTF7 alone as a negative control , and washed . The infected B-LCLs were incubated for defined time periods ( 0 , 2 , 4 , 8 , 12 , 16 , 20 and 24 hours ) at 37°C and added as stimulators to peptide-specific CTL lines prior to intracellular IFN-γ staining . For proteasome inhibition assays B-LCLs were incubated for 1 . 5 hours with 2 nM epoxomicin ( Calbiochem ) or 50 nM lactacystin ( Calbiochem ) prior to vaccinia virus infection . Procedures were performed as described previously [45] . In brief , peptide-specific CTL lines were stimulated with HLA-matched B-LCLs pulsed with increasing concentrations of corresponding peptide ( 10−9–10−4 M ) , or with HLA-matched B-LCLs infected with vHCV 827 , at an E∶T ratio of 1∶1 . Cells were then incubated for 5 hours in the presence of 50 U/mL human rIL-2 and 1 µL/mL brefeldin A ( BD PharMingen ) . After incubation , cells were blocked with immunoglobulin G1 ( IgG1 ) and stained with anti-CD8-PE mAb ( BD Biosciences ) . Following fixation/permeabilization with Cytofix/Cytoperm ( BD PharMingen ) , cells were stained with anti-IFN-γ-FITC mAb ( BD Biosciences ) , then fixed in 100 µL CellFIX ( BD PharMingen ) prior to flow cytometric analysis . For multicytokine staining cells were incubated for 5 hours in the presence of 10−5 M specific peptide , 50 U/mL human rIL-2 , 0 . 325 µL/mL monensin ( BD PharMingen ) and 0 . 5 µL/mL brefeldin A ( BD PharMingen ) . After incubation , cells were blocked with IgG1 and stained with anti-CD8-APC H7 ( BD PharMingen ) mAb . Following fixation/permeabilization with Cytofix/Cytoperm , cells were stained with anti-CD107a-PE ( BD PharMingen ) , anti-IFN-γ-eFlour450 ( eBioscience ) , anti-IL-2-PerCP-Cy5 . 5 ( BioLegend ) , anti-MIP-1β-FITC ( R&D Systems ) , and anti-TNF-α-PE-Cy7 ( BD PharMingen ) mAbs , or anti-IFN-γ-eFlour450 , anti-IL-4-PE ( BD PharMingen ) and anti-IL-10-APC ( BD PharMingen ) mAbs or anti-IFN-γ-eFlour450 , anti-IL-17A-PE ( eBioscience ) and anit-IL-22-APC ( R&D Systems ) mAbs then fixed in 100 µL CellFIX prior to flow cytometric analysis . HuH7-Lunet cells were transduced with the JFH1-based selectable subgenomic luciferase replicon and a selectable lentiviral vector expressing the complementary DNA of HLA-A*02 ( HuH7A2HCV ) or HLA-B*27 ( HuH7B27HCV ) [32] . Generation of the HLA-B*27 expressing vector [48] , the HLA-A*02 expressing vector [32] and the JFH1-based selectable subgenomic luciferase replicon [32] were performed as described previously . Cells were grown in Dulbecco's modified Eagle medium high glucose ( 4 . 5 g/L ) with stable glutamine ( PAA Laboratories GmbH ) supplemented with 10% fetal calf serum , 1% penicillin/streptomycin and nonessential amino acids ( Biochrom ) . For continuous passage , the culture medium was supplemented with 1 mg/mL G418 ( PAA Laboratories GmbH ) and 10 µg/mL blasticidin S hydrochloride ( Carl Roth GmbH & Co ) . HuH7A2HCV and HuH7B27HCV replicon cells were pulsed for 1 hour with increasing concentrations of corresponding peptide ( 10−9–10−4 M ) and washed intensively . Pulsed replicon cells were then cocultured with peptide-specific CTL lines at an E∶T ratio of 1∶1 . After 24 hours , the inhibition of viral replication was measured by luciferase activity . Luciferase activity was detected using the Steady-Glo Luciferase Assay System ( Promega ) and measured with Luminoskan Ascent ( Thermo Fischer ) and expressed as relative luciferase unit ( RLU ) . Constitutive and immuno-20S proteasomes were purified from LCL721 . 174 and LCL721 human EBV-transformed B cell lines as described previously [49] . LCL721 . 174 originate from LCL721 but carry only one copy of chromosome 6 , which contains a deletion of the TAP1 and 2 , LMP2 and LMP7 genes in the MHC class II region . Peptides were synthesized on a MK-IV peptide synthesizer , HPLC purified and verified by LC-MS on an HPLC Shimadzu QP8000-system ( Schafer-N ) . Additional purification to >98% purity was performed prior to proteasomal peptide digestion experiments using a JupiterProteo ( 250×2 . 1 mm ) column ( Phemomenex ) on a SMART HPLC system ( Amersham ) as described previously [29] . Proteasomal digestions were performed as described previously [29] . In brief , 2 nmol of each peptide was incubated for 2 , 4 and 6 hours with 2 µg ( or 4 µg in the degradation experiment ) of either purified immunoproteasome or constitutive proteasome in digestion buffer ( 20 mM HEPES-NaOH pH 7 . 6 , 2 mM MgAc2 , 0 . 5 mM DTT ) , and samples were diluted 1∶5 before injection into the mass spectrometer . All peptide digests were performed on the same day . Although a range of peptide fragment lengths was obtained following digestion of the two HLA-A*02-restricted epitope-containing 25-mers ( NS3: 1065AT1099 , A02-CINGVCWTV; and , NS5B: 2587VY2611 , A02-ALYDVVSKL ) , similar to results reported previously for HIV 25-mer peptides [29] , the HLA-B*27 epitope-containing peptide ( NS5B: 2833ID2857 , B27-ARMILMTHF ) was completely degraded; only 4–9 mer peptides were detected , and none contained the epitope . Consequently , the experiments were repeated with a 1∶4 dilution of the purified immunoproteasome or constitutive proteasome stock solution , and these samples were diluted 1∶2 before injection into the mass spectrometer . Mass spectrometry analyses of peptide digests were performed as described previously [29] . In brief , peptide digests were analyzed by nanoscale liquid chromatography using a Waters NanoAcquity UPLC system , with a Waters NanoEase BEH-C18 , 75 µm×15 cm reversed phase column . Mass spectrometry analysis of peptide fragments was performed using a Waters Q-Tof Premier in positive Vmode equipped with a nano-ESI . For fragment identification and relative quantification of the peptide fragments , the instrument was run in MSE-mode . Each sample was analyzed in triplicate . Data processing , fragment identification and quantification of LC-MSE data was performed using ProteinLynx Global Server ( PLGS ) version 2 . 2 or MassLynx4 . 1 software . Quantification of the remaining substrate was performed using LC-MSE data and ProteinLynx Global Server ( PLGS ) version 2 . 2 , and peptide identifications were assigned by database searches containing the full-length peptide sequences . The mass error tolerance values were mostly under 5 ppm . Precursor degradation rates were quantified manually via extracted ion chromatograms . Statistical analysis was performed using GraphPad Prism 5 software ( GraphPad Prism Software , Inc . ) . P-values were calculated using the Mann-Whitney U-test . | HLA-B*27 has a protective effect in hepatitis C virus ( HCV ) infection which could be linked to a single highly immunodominant HLA-B*27-restricted CD8+ T-cell epitope . However , the immunological mechanisms determining this protective effect are poorly understood . In this study , we analyzed multiple immunological determinants that may contribute to the protective role of the HLA-B*27-restricted HCV-specific epitope and its strong immunodominance and compared them with HLA-A*02-restricted HCV-specific epitopes . Our data indicate that the protective effect of the HLA-B*27-restricted epitope cannot be explained by a higher sensitivity for antigen stimulation , a higher proportion of effector-functions or a superior ability to inhibit viral replication of epitope-specific CD8+ T cells . We also did not find a higher naïve precursor frequency of HLA-B*27-restricted CD8+ T cells . However , we could show that the peptide region containing the HLA-B*27-restricted epitope is characterized by rapid antigen processing that was mostly due to the hydrophobic flanking regions of the epitope . This results in a faster presentation of the epitope at the cell surface of antigen presenting cells . Our results suggest that rapid antigen processing may be a key mechanism contributing to the protective effect of the immunodominant HLA-B*27-restricted epitope . These findings have clear implications for the design of antiviral vaccines . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"adaptive",
"immunity",
"immune",
"cells",
"major",
"histocompatibility",
"complex",
"antigen",
"processing",
"and",
"recognition",
"immunity",
"immune",
"activation",
"antigen-presenting",
"cells",
"t",
"cells",
"immunology",
"biology"
] | 2012 | Rapid Antigen Processing and Presentation of a Protective and Immunodominant HLA-B*27-restricted Hepatitis C Virus-specific CD8+ T-cell Epitope |
The hippocampus is the main locus of episodic memory formation and the neurons there encode the spatial map of the environment . Hippocampal place cells represent location , but their role in the learning of preferential location remains unclear . The hippocampus may encode locations independently from the stimuli and events that are associated with these locations . We have discovered a unique population code for the experience-dependent value of the context . The degree of reward-driven navigation preference highly correlates with the spatial distribution of the place fields recorded in the CA1 region of the hippocampus . We show place field clustering towards rewarded locations . Optogenetic manipulation of the ventral tegmental area demonstrates that the experience-dependent place field assembly distribution is directed by tegmental dopaminergic activity . The ability of the place cells to remap parallels the acquisition of reward context . Our findings present key evidence that the hippocampal neurons are not merely mapping the static environment but also store the concurrent context reward value , enabling episodic memory for past experience to support future adaptive behavior .
The hippocampus mediates the formation of adaptive memory for positive or negative experiences [1] , but the neurophysiological mechanisms of this learning process remain unknown [2] . The hippocampus may encode locations independently from the stimuli and events that are associated with these locations [3] . Recent findings deduced artificial association between place cells and place preference through the use of optogenetic [4–6] or electrical stimulation [7] . These results provide key evidence linking place cell activity and context-dependent encoding of space [8] . However , it remains unclear if the place cells are simply coincidence detectors or they actively mediate the learning between reward and location . To address this question , we address here 2 possibilities: if place cells don’t integrate information about location and reward , then after global remapping , the distribution of place fields should not be biased towards the location previously associated with reward . Alternatively , if place cells do integrate information about both location and reward , then after global remapping , the distribution of place fields should be precisely biased towards the location previously associated with reward . One remarkable but underexplored feature of the place cells is their ability to accumulate in locations of the environment that are consistently gainful over repeated exposure . Place fields tend to accumulate near the platform of the water maze , in which the percentage of cells with peak activity around the hidden platform was more than twice the percentage firing in equally large areas elsewhere in the arena [9] . CA1 place fields preferably map locations , such as the escape platform location in an annular water maze [9] , selective delivery of water to a single location [10] , or the food reward location in a T-maze [11] . The accumulation phenomenon has been described but it never has been validated as a learning mechanism . The biased mapping might simply reflect oversampling of a small number of place cells with no relation to the learning of the task . The place cells from the residual , nonrewarding locations of the environment may simply undergo incomplete field formation due to insufficient path sampling [12 , 13] . In this case , remapping of the place cells triggered by the altered spatial navigation approach will dissociate the accumulated place fields from the animal’s preferred location . An alternative proposal is that the accumulation of the place fields is essential for the representation of the reward location . In this case , the scale of accumulation will consistently reflect the degree of place preference , even after scattered allocation of the place fields . We use here a behavioral setup in which , after the learning trials , the place cells undergo global remapping due to the altered spatial navigation approach of the animals during the probe . We designed a protocol to allow for significantly expressed place preference in combination with sufficient path sampling for place field formation in the nonpreferred zone . Previous findings indicated that spatial learning regulates place fields accumulation [14] . Here , we present explicit evidence that the accumulation of place cells is independent population-code mediating the integration of spatial navigation and reward location . We then show that accumulation of place fields is an experience-dependent plastic process , which depends on spatially tuned tegmental dopaminergic activation .
To dissociate the place field maps from consistent reward location , we used a continuous T-maze . We trained rats implanted with tetrodes in the CA1 region of the dorsal hippocampus to navigate in a continuous T-maze task , in which the southwest ( SW ) corner was the constant reward location ( Fig 1A ) . To achieve differential navigation among the rats during the probe , we set the illumination of the recording room to levels at which the animals would rely on both distal and proximal cues ( see Materials and methods , Continuous T-maze task ) . The distal cues represent geometric signs on the curtains around the recording arena , while the proximal cues refer to the maze geometry . All rats ( n = 20 ) underwent 9 training sessions for 3 days , during which the animals learned to navigate towards the SW corner ( Fig 1D ) . During the probe session , the rats were placed in the opposite T-maze , with rewards positioned in both corners of the maze ( Fig 1B ) . The rats showed 2 types of navigation strategy ( Fig 1C ) : ( 1 ) preference for the northeast ( NE ) corner passes , which was above chance level ( S1A Fig , preference group , n = 10 ) , with binomial probability values of p < 0 . 05 ( Fig 1E , S1 Table ) ; i . e . , navigation predominantly based on proximal cues , and ( 2 ) no preference between corners in which the number of passes to each of the corners was below chance level ( S1B Fig , nonpreference group , n = 10 ) , with binomial probability values of p > 0 . 05 ( Fig 1F , S1 Table ) ; i . e . , navigation based on opposing proximal and distal cues . Only rats with stable waveforms were allowed to the probe session ( S2A–S2D Fig ) . We recorded 304 hippocampal place cells ( n = 20 rats ) and all of them underwent global remapping in the probe trial ( S2E–S2H Fig ) . From the 241 cells that fired in the reward-associated loop ( with navigation towards the SW corner ) of the maze during training sessions ( reward loop cells ) , 74 . 2% ( 179/241 ) remapped , while 25 . 8% ( 62/241 ) of place cells did not express place field for the maze configuration of the probe . Concurrently , 63 other place cells expressed place fields in the probe . These place cells were units that either expressed fields in the early training sessions of the nonrewarding loop of the maze or did not express any fields for the training maze configuration ( nonreward loop cells ) . Five out of 179 ( 2 . 79% ) of the remapped cells kept their location in respect to the maze geometry ( mirror representation ) , while 6/179 ( 3 . 35% ) kept their location in respect to the distal cues ( opposite representation ) . Our maze setup allows for a combination of place fields global remapping with concurrent preferential navigation . Importantly , the maze design allowed for sufficient path sampling in the nonpreferred section of the maze . We next examined if the remapped place fields accumulated within the preferred navigation of the probe . We investigated whether the configuration of the remapped place fields from the reward loop in the training sessions ( reward loop cells ) differed between the preference ( S3A Fig ) and nonpreference groups ( S3B Fig ) . We evaluated the spatial field configuration ( SFC ) of the individual place fields across both loops with respect to the midline of the maze at 45° , in which the SW corner is 0° and the NE corner is 90° . SFC evaluates the position of the individual place fields across the midline axis ( in degrees ) . The mean SFC of the reward loop cells for the preference group ( Fig 2A ) was 59 . 9 ± 3 . 2° , compared to 35 . 9 ± 3 . 2° for the nonpreference group ( Fig 2C ) . The mean SFC of the place fields that did not encode the reward loop in the training sessions ( nonreward loop cells ) ( Fig 2B and 2F ) was opposite to the reward loop cells fields with values of 30 . 9 ± 5 . 5° and 47 . 8 ± 5 . 7° for the preference and nonpreference group , respectively ( Fig 2D ) . The total SFC , including reward and nonreward loop cells , was 52 . 1 ± 3 . 0° for the preference and 38 . 9 ± 2 . 8° for the nonpreference group ( Fig 2E , S2 Table ) . The duration of the probe ( 12 minutes ) was designed to allow for sufficient sampling of all bins of the maze ( for field evaluation , we used a minimum of 9 bins ) , including sufficient time in the nonpreferred zone for the formation of stable place fields [13] . To confirm the clustering of the cells during the probe for the preference group of rats , we used another analytical approach . We evaluated the location of the center of mass ( COM ) and its position ( spatial angle ) in respect to the symmetry axis of the maze ( S4A–S4D Fig ) . The mean COM angle ( S4D Fig ) of the reward loop cells for the preference group was 56 . 4 ± 2 . 5° , compared to 39 . 6 ± 1 . 8° for the nonpreference group ( S4C Fig ) . The mean COM angle of the place fields that did not encode the reward loop in the training sessions ( nonreward loop cells ) was 37 . 5 ± 3 . 9° and 49 . 0 ± 3 . 8° for the preference and nonpreference groups , respectively , while the total COM angle , including reward and nonreward loop cells , was 51 . 2 ± 2 . 2° for the preference and 41 . 9 ± 1 . 7° for the nonpreference groups ( S4C Fig , S3 Table ) . These data show that the place preference behavior was accompanied by biased configuration of the reward loop cells towards the preferred NE corner of the maze . The nonreward loop cells counterpoised the SFC bias for both groups . To evaluate whether the spatial distribution of the place field assemblies reflects the navigation preference of each animal , we analyzed the COM from all place cells’ spikes recorded from a single animal using a spatial population vector ( SPV ) . This parameter estimates the Cartesian distribution of the spikes from multiple place fields . The SPV is based on place field rates , which represent spiking as a function of the occupancy for each pixel ( see Materials and methods , SPV ) . Therefore , the SPV is not biased by the time spent in a particular section of the maze . The SPV values are measured also between the SW corner , where SPV is 0° , and the NE corner , where SPV is 90° . Values below 45° indicate that the place fields distributed preferably towards the SW corner of the maze , while values of above 45° indicate NE distribution preference . We computed both weighted SPV ( in which the cells are weighted by their firing rate ) and averaged SPV ( in which all cells are weighted equally ) . The reward loop cells from the preference group ( Fig 3A and 3E , S5A Fig ) showed an uneven distribution of their spikes in favor of the NE corner , with weighted SPV of 56 . 1 ± 1 . 6° and averaged SPV of 56 . 2 ± 1 . 5° . Concurrently , the place cells from the nonpreference group ( Fig 3B and 3E , S5B Fig ) expressed values of 40 . 3 ± 2 . 0° for weighted SPV and 39 . 8 ± 1 . 6° for averaged SPV ( S4 Table ) . The addition of the nonreward loop cells shifted the SPV values towards the midline of 45° ( Fig 3F ) . The SPV values decreased for the place cells from the preference group ( Fig 3C ) to 52 . 4 ± 1 . 2° and 51 . 3 ± 1 . 4° weighted and averaged SPV , respectively ( Fig 3G , S5C Fig ) , whereas the SPV values for the nonpreference group ( Fig 3D ) increased to 42 . 7 ± 1 . 4° and 42 . 0 ± 1 . 1° , respectively ( Fig 3H , S5D Fig ) . The close link between place preference and place field assembly distribution is best represented by the correlation between the animals’ navigation and the SPV . The degree of place preference , expressed by the SW/NE passes ratio , showed strong correlation with the SPV values ( Pearson’s r = −0 . 92 , Fig 3I left ) . This correlation was not affected by the presence of the nonreward loop cells for the weighted SPV ( Pearson’s r = −0 . 90 , Fig 3I right ) . However , the averaged SPV correlation was greater for the reward loop cells ( Pearson’s r = −0 . 91 , Fig 3J left ) compared to all cells ( Pearson’s r = −0 . 75 , Fig 3J right ) . Thus , the firing rate might complement the distribution of the place cells for preferred location . To demonstrate that the correlation of the SPV and animals’ navigation is not affected by the analytical design , we forced biased navigation to the south section of the maze during the probe ( S6A Fig ) . Despite the high SW/NE passes ratio and the predominant timing in the SW corner ( S6B Fig ) , the SPV was directing towards the opposite NE corner ( S6C Fig ) . These findings provide key evidence that accumulation of place fields persists after global remapping , and the scale of place field assembly distribution precisely reflects the degree of place preference . The ventral tegmental area ( VTA ) is a central structure in the propagation of reward signals [15 , 16] . We recorded the activity of slow-spiking neurons ( with firing rate of <10 Hz , i . e . , the rate diapason of dopaminergic neurons ) from VTA and measured their firing rates for the choice points of the probe exploration ( S7A Fig ) . Cue-evoked activity in tegmental dopaminergic neurons reflects the value of the predicted rewards [17 , 18] . We evaluated separately the firing rate for the passes towards the NE corner and towards the SW corner ( S7B Fig ) . To evaluate the dissimilarity of the firing rate in both directions , we divided the firing rate for the SW passes over the firing rate for the NE passes ( SW/NE passes firing rate ratio ) . The average ratio values for animals with preferred navigation ( n = 3 rats , 16 cells ) was 0 . 58 ± 0 . 04 , compared to 1 . 01 ± 0 . 03 for animals with preferred navigation ( n = 4 rats , 14 cells ) ( S7C Fig ) . The significantly lower ratio for the preference group animals ( S7D Fig ) indicates that the tegmental slow-spiking neurons spike with higher rate when the animals from this group are navigating towards their preferred section of the maze ( S7E Fig ) . These data propose that the dopaminergic signaling might mediate the navigation-related bias of the place fields’ distribution . We next aimed to induce place preference behavior without altering the spatial navigation approach or reward location , but by suppressing the reward signals in the brain [19] . VTA dopaminergic suppression is known to evoke place avoidance [15] . Our goal was to test if dopaminergic signaling mediates the integration of location and reward encoding from hippocampal neurons . For inhibition of the VTA tyrosine hydroxylase positive ( TH+ ) neurons , we injected a Cre-inducible viral construct , adeno-associated virus AAV-EF1a-DIO-iC++-YFP , expressing light-activated chloride channels ( iC++ ) [20] , in the TH::Cre rat line ( Fig 4A ) . 90 ± 2% of neurons that expressed yellow fluorescent protein ( YFP ) also expressed TH , while 52 ± 8% of neurons that expressed TH also expressed YFP ( n = 5 rats , n = 1 , 116 TH cells , n = 665 YFP cells; n = 590 TH-YFP cells; Fig 4B , S8A–S8C Fig ) . Local delivery of blue light ( 473 nm ) suppressed the spiking of neurons infected with AAV-EF1a-DIO-iC++ ( Fig 4C ) . Of these cells , 90 . 9% ( 30/33 ) spiked with baseline frequency below 10 Hz , with average frequency of 4 . 7 ± 2 . 6 Hz . Firing rate of <10 Hz is an electrophysiological characteristic of VTA dopaminergic neurons [21] . The application of blue light triggered inhibition in 38% ( 30/78 ) of the recorded slow-spiking neurons ( S9A–S9C Fig ) and 5 . 5% ( 3/55 ) of the fast-spiking cells . To confirm that injection of AAV-iC++ mostly affected the TH+ neurons , we tested if photoinhibition would trigger place avoidance , which is a behavioral correlate of dopaminergic suppression [15] . We used a rectangular-shaped linear track because the navigation of the animals during the baseline recordings was the most evenly distributed between the opposite corners when compared to other tracks . Light delivery to VTA in the SW area of the track ( Fig 4D ) resulted in gradual avoidance of this section , with a decrease of SW/NE passes ratio to 0 . 79 ± 0 . 07 of total passes after the first and 0 . 74 ± 0 . 05 after the second session , compared to the baseline ratio of 0 . 95 ± 0 . 05 ( n = 5 rats , Fig 4E ) . To test the hypothesis that reward signals guide the spatial distribution of place field assemblies , we analyzed the effect of photoinhibition on the hippocampal place field assembly distribution using the SPV of place cells taken from ensemble recordings in the rectangular-shaped linear track ( Fig 5A ) . The weighted SPV for the YFP-iC++ group of rats ( n = 6 rats , 51 place cells ) shifted from 44 . 3 ± 1 . 3° to 50 . 0 ± 3 . 6° ( Fig 5B ) and 57 . 2 ± 4 . 4° after the first and the second photoinhibition session , respectively ( Fig 5D , S5 Table ) . The observed effect was a consequence of partial place and rate remapping ( S10A Fig ) . The changes in the weighted SPV and SW/NE passes ratio were significantly correlated ( Pearson’s r = −0 . 24 , p = 0 . 045 ) . No significant change was evident for the SPV of the control YFP group of animals ( n = 7 rats , 63 place cells , Fig 5C , S10B Fig ) between the baseline 45 . 4 ± 2 . 0° and the light delivery sessions ( 46 . 4 ± 3 . 3° and 46 . 1 ± 3 . 6°; Fig 5E; S6 Table ) . Similarly , the averaged SPV for the YFP-iC++ group shifted from 44 . 6 ± 1 . 4° to 50 . 5 ± 2 . 8° and 52 . 4 ± 4 . 5° after the first and second photoinhibition sessions , respectively ( S11A Fig ) . Forced biased navigation towards the NE corner with concurrent photoinhibition in the NE quadrant resulted in SPV with value opposed to the photoinhibition zone ( 42 . 2° ) , showing that SPV is not affected by the path sampling ( S12 Fig ) . These results provide evidence that the dopaminergic signals regulate the place fields’ assembly distribution , which is accompanied behaviorally by navigation preference . To examine how VTA projections affect hippocampal neuronal spiking , we implanted rats with an optical fiber and recording tetrodes in the pyramidal layer of dorsal hippocampal CA1 area and injected Cre-dependent AAV , which mediates blue light–induced depolarization of the dopaminergic neurons in TH::Cre rats [22] . The injection of AAV5-EF1a-DIO-ChR2-E123T/T159C resulted in specific expression of light-activated channelrhodopsin 2 ( ChR2 ) tagged with a fluorescent protein in TH+ neurons ( Fig 6A ) . Blue light delivery entrained the firing of slow-spiking neurons in the lateral VTA ( Fig 6B ) . We evaluated the effect of the light delivery on the spiking of hippocampal place cells during a pellet-chasing task in open arena . The photostimulation ( 473 nm , 50 Hz , 12 pulses , 5 ms pulse duration ) was applied every 6 seconds , including intrafield and extrafield passes . We investigated the firing frequency of 22 place cells from 3 rats during the photostimulation protocol with a duration of 250 ms as well as the neuronal firing in the first 100 ms after the protocol onset ( Fig 6C ) . The spiking of the place cells increased to 123 . 8 ± 6 . 3% of the prestimulation firing rate for the first 100 ms and 112 . 3 ± 6 . 7% for the entire protocol of 250 ms ( Fig 6D , S7 Table ) . The photostimulation effect was mediated by the intrafield spike rate increase , whereas the light delivery did not affect the number of extrafield spikes ( S13A–S13C Fig , S8 Table ) . Concurrently , the photostimulation reduced the firing rate of 21 slow-spiking interneurons ( cells with firing rate <10 Hz , Fig 6E ) to 86 . 5 ± 4 . 4% for the first 100 ms and 78 . 4 ± 2 . 7% for 250 ms ( Fig 6G , S7 Table ) . We identified a functional connection between the slow-spiking interneurons and the place cells ( Fig 6F , S14A–S14C Fig ) . The spike cross correlation indicates a monosynaptic connection between cell pairs [23 , 24] . The position of the cross correlation peak in relation to time 0 indicated that the place cells in our recordings were presynaptic , while the slow-spiking interneurons were the postsynaptic neuron of each pair . These data show that dopamine signal enhances the excitability of the hippocampal place cells , whereas for a subset of postsynaptic slow-spiking interneurons , dopaminergic signaling gradually reduces their ability to trigger spikes . To evaluate the causality of VTA activation on the direction of place field center remapping , we applied photostimulation tangential to the place fields recorded in the open field arena during a pellet-chasing task . With the open arena , we have eliminated the goal-directed [11] and directional place field plasticity [25] occurring in linear tracks with prospective reward location . We photostimulated the dopaminergic fibers in hippocampal CA1 of TH-Cre rats injected in VTA with AVV-ChR2-YFP ( Fig 7A ) and evaluated the field properties of hippocampal place cells ( S1 Data ) . We evaluated the distribution of place cell spikes across the subsequent recordings: baseline ( Fig 7B ) , first photostimulation ( Fig 7C ) , second baseline ( Fig 7D ) , and second photostimulation ( Fig 7E ) . We estimated if there is a shift in the place fields’ COM and measured the distance ( Fig 7F ) . We compared the field properties of place cells of the TH-Cre rats injected with AVV-ChR2-YFP ( ChR2 group ) to the cells from animals injected with control viral vector ( YFP group , S15A and S15B Fig , S2 Data ) . We observed a gradual shift increase of the COM ( ΔCOM ) between the recording sessions ( Fig 7G ) for cells ( n = 18 ) from animals injected with AVV-ChR2-YFP ( ChR2 , n = 4 rats ) but not for cells ( n = 16 ) from control rats ( YFP , n = 3 rats ) injected with AVV-YFP . To determine if the place field shift is directed towards the location of the applied light pulses , we used a specific measure ( i . e . , Bhattacharyya distance metric [bhatt] , Fig 7H ) . Bhattacharyya distance quantifies the distance between the distribution of the place cell spikes and the distribution of the light pulses , which is constant ( lower bhatt values mean higher overlap of both distributions ) . The bhatt value in our experiments was gradually reduced after the first photostimulation , second baseline , and second photostimulation only for the cells from the ChR2 group of rats ( Fig 7I , in which the ratio of baseline over ChR2 increased ) but not for the cells from control rats ( YFP group , Fig 7I ) . There was no significant change in the peak and mean place field rate or in the spatial coherence of the place field ( S16A Fig ) . A transient increase of the field size paralleled the remapping process ( S16B Fig ) . These data show that photostimulation of the dopaminergic fibers evoked field plasticity . Furthermore , field ΔCOM shifted towards the stimulus location . We next implanted optic fiber for light delivery in VTA to evaluate if the sparse TH+ projections evoke distributed place field plasticity across the hippocampal network ( Fig 8A ) . The second baseline recording showed that the ChR2 group ( Fig 8B ) of cells shifted their COM ( ΔCOM ) 24 hours after the first photostimulation session ( 6 . 96 ± 1 . 21 cm , n = 3 rats , 17 cells , Fig 8C and 8D ) , whereas ΔCOM for the control YFP group was smaller ( 2 . 51 ± 0 . 5 cm , n = 3 rats , 16 cells , Fig 8D ) . The distribution of overlap between the photostimulated field and the place field measured by bhatt increased only for the ChR2 group ( S3 Data ) but not for the YFP controls ( Fig 8E ) . The increased overlap indicates that the direction of ΔCOM shift was tuned towards the photostimulation coordinates only for the ChR2 group . The correlation between ΔCOM and Bhattacharyya distance was significant for the ChR2 ( Pearson’s r = 0 . 36 , p = 0 . 036 , Fig 8F ) but not for the YFP group ( Pearson’s r = 0 . 05 , p = 0 . 784 ) . We also tested an alternative method for the activation of VTA projections to the hippocampal formation . Stimulation of the medial forebrain bundle ( MFB ) ( S17 Fig ) , which contains dopaminergic projections from VTA , was expected to exert a similar effect on the place field plasticity . We applied nonselective electrical MFB stimulation ( S18 Fig , S4 Data ) and confirmed a significant correlation between ΔCOM and Bhattacharyya distance ( Pearson’s r = 0 . 37 , p = 0 . 024 , S19A Fig ) . The increase of ΔCOM was significant after the second MFB stimulation session ( 4 . 72 ± 2 . 94 cm , n = 5 rats , 18 cells , S19B Fig ) compared to the YFP controls . Together , these results indicate that place cells remap their fields in a direction determined by repeatedly augmented VTA activity .
The spatial environment can be associated with positive or negative context and repetitive exposure to such environment drives goal-directed behavior [26 , 27] . The spatial memory in rodents is behaviorally measured by their navigation in tasks where the rodents learn to associate reward or aversion with a particular location in the environment [21] . The resulting place preference is evaluated either by the path of the animals towards the preferred destination or by the time spent there [15 , 16] . One of the main challenges when investigating place cell activity during place preference tasks is the insufficient path or time spent in the nonpreferred location . Insufficient path sampling impedes the formation and evaluation of the place fields [12] . To achieve both place preference and sufficient path sampling in the nonpreferred location , we used a continuous version of the cross-maze task [28] . This experimental design allowed us to evaluate the behavior of the animals at both choice points of the maze . The continuous T-maze protocol did not induce differences in the task demands , and the consistency of the reward was independent of the animals’ behavior . The continuous T-maze design was set for the animals to rely on different strategies based on local and distal cues . In this way , the animals expressed differential navigation preference . The rats with significant place preference navigation during the probe navigated in this manner not because of the reward location ( equivalent reward was positioned in both corners ) but because of the integration of the spatial cues and reward during the training sessions . T-maze and plus-maze are behavioral setups for global remapping and this is described as journey-dependent mapping [29] . For this reason , we have chosen T-maze maze protocol and , subsequently , we have confirmed the global remapping with our data . The cross correlation of hippocampal pairs indicated global remapping during the probe; however , the distribution of the place cells for some animals was spatially biased , suggesting that the remapping was not random . Our goal here was to show that accumulation of place fields relates to the degree of preferential navigation , even after scattered remapping of the fields . We showed that place cells from animals preferably navigating in the east loop of the maze expressed a higher degree of accumulation , represented by the SFC values . A key finding of our study shows that place field accumulation encodes location-specific reward valence independently from the spatial representation code , which is reset after the global remapping . This finding supports the hypothesis that the field accumulation reflects the reward location , corresponding to the degree of place preference . Spatial configuration of the place fields in different subregions of a large environment is related to geometric or contextual similarities of these subregions [30] . The place fields recorded from the preference group of rats were characterized with more asymmetric configuration , suggesting memory-mediated contextual difference between both loops of the maze . Our findings demonstrate that the spatial configuration of the place fields varies as a function of the experience but not the actual reward location ( equivalent reward was presented in both maze corners during the probe ) . This finding suggests that the place cells from the reward loop of the training sessions integrated spatial and reward information on a population level within a functional engram [4 , 6] . Thus , after the global remapping during the probe session , the reward loop cells were distributed predominantly in the section of the maze associated with higher experience-dependent reward value . To compensate for this spatial representation bias , the nonreward loop cells were distributed predominantly in the opposite loop of the maze during the probe . Our data support the proposal that place fields encode not only the spatial geometry but also the reward expectance across the environment [31] . The study of population dynamics in hippocampal neurons is one of the most powerful tools for understanding the link between place fields and navigation [32 , 33] . To examine the proposal that fields distribution relates to preferred location , we used a metric that evaluates the distribution of the place cells within the Cartesian plane , i . e . , SPV . Weighted dimensional representation of the population dynamics is a new analytical technique in place field data analysis; however , it is already applied in the decoding of movement direction from motor cortex neuronal ensembles [34] . Using the common computational approach , here we propose that neuronal populations in different brain circuits share fundamentally similar mechanisms of information encoding and retrieval . The benefit of the SPV is that it can relate navigation to the change of assembly field distribution for different forms of remapping such as rate and place remapping for different experimental conditions and behavioral setups . Previous studies have investigated place cell spiking for individual journeys of plus-maze tasks [35–37] . Our experimental design included continuous navigation , which allowed for the estimation of place preference ( measured after the evaluation of repeated behavior ) in parallel with the formation of stable place fields ( measured after the evaluation of repeated path sampling ) . The effect of egocentric inputs was reduced to minimum because the animals were allowed to navigate in opposing directions and the behavior at the choice points relied exclusively on allocentric signals . Thus , the directionality of the place fields did not affect the SPV values . To relate the degree of SPV to scale of preferred navigation , we induced variability of the place preference behavior among the animals . The dissociation of the distal and proximal cues [38] during the probe resulted in differential path navigation strategies with sufficient path sampling of both loops . Here , using assembly measures of the remapped place fields , we established a link between the spatial distribution of place fields’ ensemble activity and the animal’s preferred navigation . The firing rate of individual place cells contributes to the SPV , indicating why weighted SPV correlates better with the navigation preference when compared to the averaged SPV . The SPV did not depend on the path sampling because the firing rate accounted for the occupancy of each pixel . The data provide evidence that the hippocampal place field assembly code closely reflects experience-dependent navigation preference . Our findings show that hippocampal neurons are not merely mapping the environment but their spatial distribution encodes learning-adopted location of a prospective reward . To understand the underlying mechanisms of reward-dependent changes in place cells’ activity , we optogenetically manipulated their dopaminergic inputs . Optogenetic stimulation of hippocampal dopaminergic fibers arising from VTA in mice during spatial learning of novel locations improves the place field stability and stabilizes spatial memory performance [21] . An alternative way to show the importance of dopaminergic signaling to hippocampal spatial representation is suppression of the VTA-generated signal . Accordingly , we found that photoinhibition of dopaminergic VTA activity evoked reconfiguration of the place fields . We used a novel adeno-associated virus iC++ , which triggered chloride-conducting photoinhibition after blue light application [20] . The suppression of VTA dopaminergic neurons evoked partial place field remapping and shifted the SPV towards the arm without photoinhibition . Our data provide direct experimental evidence of the relationship between the hippocampal assembly configuration and the activity of midbrain dopaminergic neurons , encoding salient information [39 , 40] . The hippocampal neurons encode newly learned goal locations through the reorganization of assembly firing patterns in the CA1 region [33 , 41] and this process is NMDA-receptor dependent [14] . Our findings highlight the important role of dopaminergic signaling in field assembly reorganization . Moreover , dopamine-dependent shifts in the SPV indicated the direction of behavioral navigation preference . Our results extend previous findings that knockout mice for dopamine receptor D1 receptor do not remap in response to environmental manipulation [42] . A major advantage of the use of rats for behavioral optogenetic experiments is the high degree of path sampling in achieved continuous navigation experiments . This degree of path sampling is essential for the precise evaluation of the place field parameters , including the COM . We implanted transgenic TH::Cre rats with optic fibers in the lateral section of the VTA , a region with a high degree of segregation between the TH and GAD expressing neurons [43] . In addition , the expression of adeno-associated virus in TH cells shows varying degrees of selectivity and penetrance; for our expression , the selectivity was 52% for AAV-iC++ . This degree of expression was sufficient for the behavioral response of laser light application in our rectangular-shaped linear maze . The photoinhibition successfully induced place avoidance after 2 sessions . This finding supports the proposal that midbrain dopaminergic inputs are central to the integration of salience and hippocampus-dependent memory in rodents [39 , 40 , 44] . Functional magnetic resonance imaging ( fMRI ) studies in humans also confirm that VTA activity correlates with enhanced learning in the context of novelty [45] . Optogenetic manipulation of VTA cells allowed us to regulate the reward inputs to the entire dorsal hippocampal network , compared to local stimulation of dopaminergic fibers . Activation of the sparsely connected dopaminergic projections can induce a sufficient effect on the hippocampal spatial network to reflect navigation preference . The system effect of VTA activity on behavior may involve also the ventral striatal neurons that generate firing patterns correlating to task events such as prospective rewards , goal locations , and sensory stimuli predicting rewards [46 , 47] . Another source of dopaminergic regulation of hippocampal spatial representation may originate from locus coeruleus . Dopamine coreleasing TH+ neurons in locus coeruleus mediate postencoding memory enhancement and optogenetic activation of these cells evokes novelty-associated memory enhancement [48] . These recent findings propose that the dopaminergic innervation in the hippocampus may need to be further examined , with particular focus on the locus coeruleus . Locus coeruleus could exert a more potent effect on hippocampal physiology and spatial representation compared to VTA in particular tasks involving novelty exploration and memory formation [48] . The photostimulation of dopaminergic projections in the hippocampus after AAV-ChR2-YFP virus injection in VTA resulted in increased spiking of the place cells to 123 . 8% , while slow-spiking interneurons ( with monosynaptic inputs from the place cells ) gradually suppressed their firing to 78 . 4% throughout the light delivery protocol . The recorded slow-spiking interneurons in the CA1 pyramidal cell layer share characteristics similar to those of Ivy cells , which are interneurons expressing neuropeptide Y ( NPY ) or the neuronal nitric oxide ( NO ) synthase isoform [49] . Parvalbumin-expressing interneurons in the hippocampal CA1 pyramidal cell layer ( basket , bistratified , and axo-axonic cells ) often display a faster spiking pattern [50] . The firing rate of the Ivy cells recorded in behaving animals is 2 . 4 ± 1 . 8 Hz during theta episodes and 3 . 0 ± 3 . 6 Hz during non-theta episodes [49] . In our recordings , the average spiking of this group of interneurons was 3 . 29 ± 2 . 3 Hz . Paired recordings in vitro showed that Ivy cells receive depressing excitatory postsynaptic potentials from the pyramidal cells [49] . Similarly , we found that photostimulation-augmented place cell activity was paralleled by gradually suppressed firing of postsynaptic slow-spiking interneurons . Reciprocally , Ivy cells regulate the excitability of pyramidal cell dendrites through slowly rising and decaying GABAergic inputs [49] . Increased firing rate of place cells through disinhibition has been recently proposed as one of the hippocampal mechanisms for rate remapping [51] . A similar mechanism might mediate the rate remapping observed in our recordings , while the field remapping may reflect a dopamine-specific plasticity mechanism . The observed suppression of hippocampal slow-spiking interneurons might facilitate the experience-dependent plasticity effect of dopaminergic inputs on the place cells . Concurrently , we show that optogenetic activation of the dopaminergic VTA neurons evoked a gradual shift of the COM of the place fields when the dopaminergic neurons were activated in close proximity to the place field . Furthermore , the bias of the observed ΔCOM indicated the spatial direction of dopaminergic photostimulation . Thus , the remapping of COM was evoked by location-specific dopaminergic activation . The direction of place field plasticity was evaluated by the Bhattacharyya distance overlap . The observed reduction of bhatt values indicated ΔCOM was due to displacement in the direction of the stimulation coordinates . Our data show dopamine-dependent place field plasticity , which may be the spatial substrate of dopamine-mediated long-term synaptic plasticity . Blockade of dopamine D1/D5 receptors in CA1 impairs the long-term synaptic plasticity in vivo [52] . Furthermore , novelty-induced release of dopamine in the hippocampus [53] facilitates long-term plasticity , which is prevented by blocking of D1/D5 receptors [54] . Finally , long-term synaptic plasticity is believed to be the cellular mechanism of learning and memory [55] . In summary , our data show that place cells do not passively map the spatial environment in a Cartesian coordinate system but continually tune their fields to encode the location of prospective reward on population level . The place fields’ spatial plasticity is the key element in their ability to undergo assembly redistribution in order to mediate memory formation .
We conducted our experiments in accordance with directive 2010/63/EU of the European Parliament , the council of 22 September 2010 on the protection of animals used for scientific purposes , and the S . I . No . 543 of 2012 and followed the Bioresources Ethics Committee , Trinity College Dublin , Ireland ( individual authorization number AE19136/I037; procedure numbers 230113–1001 , 230113–1002 , 230113–1003 , 230113–1004 , and 230113–1005 ) , and international guidelines of good practice under the supervision of Marian Tsanov , who is licensed by the Irish Medical Board ( project authorization number: AE19136/P003 ) . Lister hooded rats have been chosen for these experiments because of their anatomical and physiological similarities to humans . Several steps were taken to minimize stress in the animals , which helped during surgery and recovery . Animals were handled and allowed to grow accustomed to their environment well before surgery . The surgery was completed as quickly and safely as possible to reduce the recovery time . The surgery itself was undertaken in aseptic conditions to reduce the risk of infection and anesthesia was carefully monitored to ensure that the animal was not stressed or in pain . After the surgery , a painkiller and antibacterial were administered to aid in recovery . We undertook to comply with all the ethical and security issues by appropriate protocols to ensure the application of the 3 R’s ( reduction , replacement , and refinement ) . Male , 3–6-month-old , Lister hooded TH::Cre rats ( Rat Resource & Research Center P40OD011062 , United States ) were individually housed for at least 7 days before all experiments under a 12-h light–dark cycle . The animals accessed water ad libitum . Rats were food deprived to 80% of their original weight . The recording sessions were conducted during the light phase . Eight tetrodes were implanted in the hippocampal CA1 area: −3 . 8 AP , 2 . 3 ML , and 1 . 8 mm dorsoventral to dura . For recordings of dopaminergic neurons , 8 tetrodes were implanted unilaterally in VTA: 5 . 7 AP , 1 . 9 ML , angle 10o medially , and 8 . 0 mm dorsoventral to dura . We targeted the lateral VTA due to greater colocalization of TH and targeted dopaminergic neurons in the TH-Cre rodent lines compared to the medial VTA [56] . The recordings were performed as previously described [57] . After a minimum 1-week recovery , subjects were connected via a 32-channel headstage ( Axona Ltd . ) to a recording system , which allowed simultaneous animal position tracking . Signals were amplified ( 10 , 000–30 , 000-fold ) and band-pass filtered between 380 Hz and 6 kHz for single-unit detection . To maximize cell separation , only waveforms of sufficient amplitude ( at least 3 times the noise threshold ) were recorded . Candidate waveforms were discriminated offline using graphical software ( Tint , Axona Limited ) , which allows waveform separation based on multiple features including spike amplitude , spike duration , maximum and minimum spike voltage , and the time of occurrence of maximum and minimum spike voltages . To verify that spike sorting analysis targets the recording from individual cells , we examined the following parameters for each unit: ( 1 ) waveform , the spike from a single neuron is characterized with consistent amplitude and time of occurrence of maximum and minimum spike voltages , ( 2 ) spike cluster , the spike cluster of an individual spike has a characteristic shape and it is located at consistent coordinates on the cross-amplitude scatterplots , ( 3 ) spike autocorrelogram , the first 2 ms represent the refractory period . Autocorrelation histograms were calculated for each unit , and the unit was removed from further analysis if the histogram presented spiking within the first 2 ms ( refractory period ) , inconsistent with good unit isolation . Only stable recordings across consecutive days were further analyzed . The stability of the signal was evaluated by the cross correlation of spike amplitudes and similarity comparison of the spike clusters between the training sessions . Electrode stability was assessed offline by comparison of waveforms and cluster distributions . The single-unit signals of the last recording session and the probe were compared for waveform similarity , cluster location , size , and boundaries . Peak and trough amplitudes of the averaged spike waveforms were compared by Pearson’s r [36] . r values ≥ 0 . 9 indicated that the same populations of cells were recorded throughout the last recording session and the probe . We analyzed the single-unit recordings as previously described [57] . Single hippocampal pyramidal cells and interneurons were identified using spike shape and firing frequency characteristics [58] . Firing rate maps allow for visual inspection of neurons’ preferred areas of firing ( i . e . , place fields ) . They were constructed by normalizing the number of spikes that occurred in specific pixelated coordinates by the total trial time the animal spent in that pixel . This produced maps depicting the place fields of each cell and quantified in Hz ( smoothed maps ) . Place field was defined as the region of the arena consisting of at least 9 adjacent bins , which contain spikes . We defined place field size as the region of the arena in which the firing rate of the place cell was 20% or greater of the maximum firing frequency [59] . We used multiple measures to analyze the spatial properties of the hippocampal place cell firing ( i . e . , place field size , spatial selectivity , spatial coherence , and spatial specificity information content ) . The spatial information of a firing field ( ratio of maximal signal to noise ) was calculated by dividing the firing rate of the cell in the bin with the maximum average rate by its mean firing over the entire apparatus . Spatial coherence consists of a spatial autocorrelation of the place field map and measures the extent to which the firing rate in a particular bin is predicted by the average rate of the 8 surrounding bins . Thus , high positive values result if the rate for each bin can be better predicted from the firing frequency of a neighboring location . With each spatial autocorrelation performed on the place field map , a p value is calculated , indicating whether the correlation is statistically significant . Place field activity is not considered to be spatially coherent if the p value is greater than 0 . 001 . The spatial information ( or spatial specificity ) is expressed in bits per spike and is calculated as follows I=∑iPi ( λi/λ ) log2 ( λi/λ ) where λi is the mean firing rate in bin i , λ is the overall mean firing rate , and Pi is the occupancy probability of bin i . The spatial specificity index is a measure of the amount of information about the location of the animal conveyed by a single spike generated by a single place cell . Global remapping indicates relocation of the place field for each of the recorded place cells . Due to the binning of the T-maze ( 10 cm width , 85 cm length of the leg arms ) for place field analysis ( 2 . 5 cm/bin ) , the chance of a place field composed of ≥3 x 3 bins to appear on the same location during a random global remapping is 1/43 ( 2 . 32% ) . Consistent with this prediction , 5/179 ( 2 . 79% ) of the remapped cells kept their location in respect to the maze geometry ( mirror representation ) , while 6/179 ( 3 . 35% ) kept their location in respect to the distal cues ( opposite representation ) . The global remapping was evaluated by the cross correlation of pairs of cells between the training session and the probe , which represents the degree of spatial overlap between the place cells . Color-coded cross-correlation matrices visualize the Spearman’s correlation between each pair of cells . Full overlap of the maps is denoted with red and a value of 1; no overlap is denoted with green and a value of 0 , and a spatially inversed map is denoted with blue and a value of −1 . For statistical analyses , the correlations were transformed into z values . The continuous T-maze task evoked a reward-driven place preference , the location of which was associated with both maze geometry ( source of proximal cues ) and distal allocentric cues . The identification of the distal cues was designed in a manner in which approximately half of the animals would rely on them for spatial navigation ( see below ) . During the training sessions , the animals were placed on the starting choice point and allowed to explore the maze ( 10 cm width , 85 cm length of the leg arms ) . The access to the opposite T-maze was disconnected . Two pellets ( TestDiet , Formula 5TUL ) were continuously positioned at the end of the SW corner ( reward zone ) , while no pellets were positioned in the NE corner . In 33% of the passes , 1 pellet was positioned at the central choice point and in 33% of the passes , 1 pellet was positioned at the starting choice point . In the remaining 33% of the passes , no pellet was present at the choice points . The sequence of pellet positioning in the choice points was random . In this way , the animals adopted continuous locomotion in a loop pattern through the choice points . Direct passes between the SW and NE zones were not rewarded with pellets . The animals were allowed to navigate in both clockwise and anticlockwise directions of both arms . As such , the animals were not able to rely only on egocentric learning , as the central and the starting choice points required opposite head turns . Thus , the animals were required to learn to navigate with reference to proximal cues ( choice point shapes ) and/or distal cues ( white plus , grey star , blue square , and yellow circle signs attached on the black curtains surrounding the maze ) . The duration of each trial was 12 minutes . Rats were given 3 daily training trials over 3 consecutive days ( 9 trials in total ) . On Day 4 , the animals were given the probe trial . During the probe trial , the animals were placed on the starting choice point of the opposite T-maze , while access to the training T-maze was disconnected . In this way , the rats were exposed to the same proximal but opposite distal cues . Two pellets were constantly positioned in both the SW and NE corners . The training and the probe recordings took place in a room with 4 distal cues ( size A4 ) and luminosity of 10–15 lux . The luminosity was set at a level such that there was approximately 50% probability that the rats would rely on allocentric navigation , dependent on distal cues [28] . The luminosity level determines the navigation strategy of the animals in the probe session of the continuous T-maze task . High luminosity ( >20 lux ) allowed the rats to identify the distal visual cues and use them as a spatial reference . During the probe , the conflict between the distal and proximal ( choice point shapes ) cues results in a split navigation strategy towards both arms . Low luminosity ( <5 lux ) results in navigation guided predominantly on the proximal maze cues , as the distal cues are an insufficient reference for spatial orientation . The task was designed to distinguish between allocentrically guided preferential and nonpreferential navigation , with a sufficient number of passes in the nonpreferential section of the maze . An insufficient number of passes results in incomplete experience for the formation of place fields and invalidates the evaluation of their properties [12] . Hippocampal neurons require 5–6 minutes of experience to form a stable spatial representation in a novel environment [13] . The chance level of preferential versus nonpreferential navigation was based on the number of passes from the choice points towards each of the corners , given the total number of passes . The binominal probability mass function was calculated as follows f ( x;n , p ) =n ! x ! ( n−x ) ! px ( 1−p ) n−x where x is the number of passes from the choice points towards 1 direction ( south or east loops ) , n is the number of passes , and p is the probability of a pass outcome towards a certain direction . SFC for continuous T-maze was based on a smoothed rate map with the size [M x M] bins and the corresponding place field map containing N place subfields . The resulting map is grouped to 6 levels ( >0 , >1/6 fmax , , >2/6 fmax , >3/6 fmax , >4/6 fmax , >5/6 fmax ) according to the maximum firing frequency of the rate map ( fmax ) and separated into leveled place fields . The maps are binary . For each level l {l ∈ ℕ| 2 ≤ l ≤ 6} an [N x N] relation matrix is created by the overlap between each map ( mn , l ) and each transposed map ( mm , kt ) of the same or lower level and the transposed place field map ( pfpt ) Rl=[r1 , 1 , lr1 , m , l⋯r1 , N , lrn , 1 , lrn , m , l⋯rn , N , l⋮⋮⋱⋮rN , 1 , lrN , m , l⋯rN , N , l] The [N x N]-sized relation matrix displays the relations between place field n at level l and place field m: rn , m , l=An , l−1∙on , m , l The relation between place field n at level l and place field m ( rn , m , l ) is the overlap between place field n at level l and place field m ( on , m , l ) , normalized by the area of place field n at level l ( An , l ) . The overlap between place field n at level l and place field m is the sum of the product of the value of bin ( i , j ) of place field n at level l ( mn , l ( i , j ) ) and the value of the bin ( i , j ) transposed place field map ( pfmt ( i , j ) ) and the value of the bins ( i , j ) transposed maps of place field m at all levels equal and smaller level ( mm , kt ( i , j ) ) . An , l=l∙∑i=1M∑j=1Mmn , l ( i , j ) where An , l is the area , expressed by the number of bins , of place field n at level l weighted by the level l . From the relation matrix , we calculate the SFC ( SFCl ) for each level by summing the rows of the relation matrices and weighting these sums by the ratio of the area of place field n at level l to the sum of the areas of all place fields in level l ( wnl ) . The place SFC of the firing characteristics of the cell is the average of the field distributions of each level weighted by the ratio of the area of all place fields in level l to the sum of the area of all place fields in all levels ( Atotal ) SFCspat=∑l=26SFCl∙∑n=1NAn , lAtotal Atotal=∑l=26∑n=1NAn , l where SFCspat is the SFC value . The COM of the place cells’ spike distribution is calculated as follows COMx=∑i=1Nx∑j=1Nyfij∙i∑i=1Nx∑j=1Nyfij∙lbin−lbin2 COMy=∑i=1Nx∑j=1Nyfij∙j∑i=1Nx∑j=1Nyfij∙lbin−lbin2 where Nx , Ny define the number of bins in the arena in X- , Y- direction; fi , j is firing frequency in bin i , j; and lbin is the bin size . Given the origin O ( Ox , Oy ) , which denotes the NW corner of the Cartesian coordinate system , and the direction of the symmetry axis D ( Dx , Dy ) , which denotes the line between the SW and NE corners , the distance of the COM ( COMx , COMy ) to the symmetry axis is calculated as follows distCOM/sym=|det[Dx−OxDy−OyCOMx−OxCOMy−Oy]| ( Px−Ox ) 2+ ( Dy−Oy ) 2 where P is the shortest distance between the COM and the symmetry line . For the rectangular-shaped linear track , the arena borders are defined as the square surrounding all motion-tracking sample points with an equal distance to the real limits of the arena at all sides . For the continuous T-maze track , the borders are defined as the hypotenuse of the arena , which is congruent with the diagonal connecting the SW and the NE corners . Using distCOM , we calculated the distance between O and P as follows: OP¯= ( COMx2+COMy2 ) −distCOM2 The COM distance normalized by the arena width perpendicular to the symmetry axis through the COM is distnorm={distCOMOP¯∙C , OP¯<OM¯2distCOM ( OM¯−OP¯ ) ∙C , OP¯>OM¯2} where OM¯ is the diagonal of a square enclosing all motion-tracking data points and C is a motion-tracking data factor set to 0 . 95 for the continuous T-maze and 0 . 85 for the linear rectangular track . The COM angle is then calculated as follows θCOM={45°∙ ( 1−distnorm ) , COMx<COMy45°∙ ( 1+distnorm ) , COMx>COMy} Using an SFC angle θSFC , we set a numerical value for the SFC , SFCspat ( 0: no symmetry , 1: maximum symmetry ) , and the location of the COM: θSFC={SFCspat*45° , COMx<COMy90°−SFCspat*45° , COMx>COMy} SFCspat: SFC value; COMx , COMy: X- , Y- coordinate of COM We used the population vector of the place field distribution D as well as the grand rate population vector F to describe the common behavior of the whole population of place cells recorded . The population vector of place field distribution consists of the COM angle of each cell θCOM , n , where n indicates the cell index Dn=θCOM , n Fn=snΔttot where Sn is number of spikes of cell n; Δttot is duration of the measurement . The SPV weighted by the mean firing frequency of the cells SPVweighted is calculated as the dot product of the grand rate population vector F and the population vector of place field distribution D normalized by the 1-norm of the grand rate population vector SPVweighted=F∙D‖F‖1 and the average SPV , SPVavg , is defined as the 1-norm of the population vector of place field distribution D , normalized by the number of place cells N in the population SPVavg=‖D‖1N The 1-norm of a vector is defined as the sum of its elements ‖F‖1=∑n=1N|Fn|;‖D‖1=∑n=1N|Dn| All algorithms were created using MATLAB . We used a Cre-inducible viral construct designed for optogenetic purposes [22 , 60] . pAAV5-Ef1a-DIO-hChR2 ( E123T/T159C ) -EYFP-WPRE-pA was serotyped with AAV5 coat proteins and packaged by Vector Core at the University of North Carolina . Viral titers ranged from 1 . 5–8 x 1012 particles per mL [22] . AAV8-EF1a-DIO-iC++-TS-EYFP was serotyped with AAV8 coat proteins , in a titer of 4 . 3 x 1012 particles per mL , provided by Karl Deisseroth ( Stanford University ) . For control experiments , we used virus bearing only the YFP reporter [22] . Randomization of group allocation ( iC++ or E123T/T159C versus YFP controls ) was performed using an online randomization algorithm ( http://www . randomization . com/ ) . The virus injection was applied unilaterally in the VTA ( 5 . 7 AP , 1 . 9 ML , angle 10° medially ) , with volume of 2 μl injected on 2 levels: 1 μl at 8 . 0 mm and 1 μl at 9 . 0 mm dorsoventral to the dura . Subsequently , an optical fiber ( 200-μm core diameter , Thorlabs , Incorporated ) was chronically inserted ( 5 . 7 AP , 1 . 9 ML , 8 . 0 DV , angle 10° medially ) . Simultaneous optical stimulation of the VTA and extracellular recording from CA1 were performed in freely behaving rats 3 weeks after the surgery . For the concurrent recordings in the hippocampal CA1 region , the optical fiber was inserted inside the microdrive cannula ( Axona , Limited ) of the recording tetrodes , with the tip of the tetrodes projecting beyond the fiber by 500 μm , and the optical fiber was coupled to a 473-nm laser ( Thorlabs , Incorporated ) . The light power was controlled to be 10–15 mW at the fiber tip . Square-wave pulses with duration of 5 ms were delivered at frequency of 50 Hz . The animals were trained to navigate between the SW and NE corners , where 2 pellets were continuously positioned . The animals were allowed to navigate in both clockwise and anticlockwise directions of the rectangular-shaped linear track ( 10 cm width , 85 cm length of the arms ) . For the optic stimulation sessions , the laser was switched on when the animal entered the north arm or the west arm , with continuous photostimulation trains ( 473 nm , 50 Hz , 5 ms pulse duration , 12 pulses per train , 0 . 5 sec intertrain interval ) until the animal exited this section of the track . The duration of each session was 12 minutes . Because the detection of our dependent variable ( navigation preference ) was independent of the experimenter , we did not use a blinding process for group allocation or behavior scoring . For the open field recordings , the rats were placed in a square arena ( 60 x 60 cm ) and 20-mg food pellets were thrown in every 20 seconds to random locations within the open field ( pellet-chasing task ) ; in this way , animals locomoted continuously , allowing for complete sampling of the environment . For optogenetic stimulation of dopaminergic fibers in the hippocampus , we applied photostimulation ( 473 nm , 50 Hz , 12 pulses , 5 ms pulse duration ) every 6 seconds during a recording session with duration of 12 minutes . We evaluated the pre- and poststimulation firing rates for 100 and 250 ms before and after the onset of the blue light train . For optogenetic stimulation of dopaminergic VTA neurons , a single train of photostimulation ( 473 nm , 50 Hz , 12 pulses , 5 ms pulse duration ) was applied when the animal crossed the borders of the selected quadrant . The experiment consisted of 4 recording sessions: first a baseline , followed by a first photostimulation session; 24 hours later , a second baseline ( baseline 2 ) was followed by a second photostimulation session ( ChR2 2 ) . The blue laser was synchronized with the video tracking and with the recoding system through DACQBASIC scripts ( Axona . Limited ) . The chosen duration of 12 minutes allowed the rats to explore evenly the arena in 2 subsequent recordings per day ( baseline and stimulation sessions ) for 2 consecutive days . Baseline recordings >15 minutes result in insufficient exploration of the subsequent stimulation session , while recordings <10 minutes reduced the sampling of the explored environment . The rats were habituated to the square arena before the recordings . Bhatt is a parameter that quantifies the distance between 2 equally sized probability distributions [61] , in which a complete overlap of 2 identical distributions is 0 bhatt . Bhattacharyya distance values are unaffected by the scale of the distributions . The Bhattacharyya distance JB between the distributions p1 and p2 is given by: JB ( p1 , p2 ) =−log∫xp1 ( x ) ∙p2 ( x ) dx We delivered electric current through electrodes ( SNEX-300 , Kopf Instruments ) implanted in the MFB: −3 . 3 AP , 1 . 8 ML , and 7 . 8 mm dorsoventral to dura . The stimulation protocol was generated by a constant current bipolar stimulus isolator ( A365D , World Precision Instruments , Incorporated ) , which was controlled through a TTL input from the recording system [57] . The protocol consisted of 4 bursts , with each burst containing 3 pulses at 10 ms ( 100 Hz ) , with an intertrain interval of 125 ms ( 8 Hz ) . The current intensities were in the range of 50–200 μA [62] and were fine-tuned individually with respect to the amplitude of the test-pulse stimulus artifact . The stimulus isolator was synchronized with the video-tracking system through DACQBASIC scripts ( Axona . Limited ) . The electrical stimulation was applied when the animal crossed the borders of the selected quadrant . The experiment consisted of 4 recording sessions: first a baseline , followed by a first MFB stimulation session; 24 hours later , a second baseline ( baseline 2 ) was followed by a second MFB stimulation session ( MFB 2 ) . At the end of the study , brains were removed for histological verification of electrode localization , as previously described [57] . Rats were deeply anesthetized with sodium pentobarbital ( 390 mg⁄kg ) and perfused transcardially with ice-cold 0 . 9% saline followed by 4% paraformaldehyde . Brains were removed , postfixed in paraformaldehyde for up to 24 hours , and cryoprotected in 25% sucrose for >48 hours . Brains were sectioned coronally at 40 μm on a freezing microtome . Primary antibody incubations were performed overnight at 4°C in PBS with BSA and Triton X-100 ( each 0 . 2% ) . The concentration for primary antibodies was anti-TH 1:200 ( Millipore , #MAB318 ) . Sections were then washed and incubated in PBS for 10 minutes and secondary antibodies were added ( 1:200 ) conjugated to Alexa Fluor 594 dye ( Invitrogen , # A11032 ) for 2 hours at room temperature . For visualization , the sections were mounted onto microscope slides in phosphate-buffered water and coverslipped with Vectashield mounting medium . The YFP fluorescence was evaluated within a selected region that was placed below the fiber tip in an area of 1 . 5 x 1 . 5 mm . Fluorescence was quantified based on the average pixel intensity within the selected region [22] . The stained sections were examined with an Olympus BX51 fluorescence microscope and an Olympus IX81 confocal microscope at 594 nm for Alexa Fluor secondary antibody and 488 nm for ChR2-YFP . TH-positive neurons were identified based on expression of red fluorescence , whereas ChR2-positive neurons were identified by expression of green fluorescence . Colocalization of Alexa Fluor 594 and YFP was determined manually using ImageJ software ( Image Processing and Analysis in Java ) . Two different approaches were used to calculate the sample size [63] . We performed power analyses to establish the required number of rats for experiments in which we had sufficient data on response variables . For experiments in which the outcome of the intervention could not be predetermined , we used a sequential stopping rule . This approach allows null-hypothesis tests to be used subsequently by analyzing the data at different experimental stages using t tests against type II error . The experiment was initiated with 4 animals per group; if the p reached a value below 0 . 05 , the testing was continued with 2 or 3 more animals to increase the statistical power . In case of p > 0 . 36 , the experiment was discontinued and the null hypothesis was accepted [63] . All data were analyzed using SPSS Software . Statistical significance was estimated by using a 2-tailed independent samples t test for nonpaired data or a paired samples Student t test for paired data . Repeated measures were evaluated with 2-way analysis of variance ( ANOVA ) paired with post hoc Bonferroni test . Correlations between datasets were determined using Pearson’s correlation coefficient . The probability level interpreted as significant was fixed at p < 0 . 05 . All data points are plotted ± SEM . | Episodic memories relate positive or negative experiences to environmental context . The neurophysiological mechanisms of this connection , however , remain unknown . Hippocampal place cells represent location , but it is unclear if they encode only the spatial representation of the environment or if they are also processing information about the reward valence for different locations . Here , we use population analysis to test the hypothesis that the place cells process the dual encoding of spatial representation and experience-dependent reward expectation . We show a unique population code for the experience-dependent value of the context . We present evidence that the accumulation of the place fields mediates the learning of the reward context of the environment . Our data reveal that the causal link between place field distribution and behavioral place preference is mediated by the tegmental dopaminergic activity . Optogenetic control of the ventral tegmental area demonstrates that dopaminergic signaling integrates the encoding of location and reward from hippocampal neurons . These findings shed a new light on the ability of hippocampal neurons to store the experience-dependent context reward value , enabling episodic memory for past experience to support future adaptive behavior . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"action",
"potentials",
"neurochemistry",
"medicine",
"and",
"health",
"sciences",
"optogenetics",
"dopaminergics",
"membrane",
"potential",
"electrophysiology",
"light",
"neuroscience",
"electromagnetic",
"radiation",
"luminescent",
"proteins",
"animal",
"behavior",
"yellow"... | 2017 | Place field assembly distribution encodes preferred locations |
Trypanosomatid protozoan parasites lack a functional heme biosynthetic pathway , so must acquire heme from the environment to survive . However , the molecular pathway responsible for heme acquisition by these organisms is unknown . Here we show that L . amazonensis LHR1 , a homolog of the C . elegans plasma membrane heme transporter HRG-4 , functions in heme transport . Tagged LHR1 localized to the plasma membrane and to endocytic compartments , in both L . amazonensis and mammalian cells . Heme deprivation in L . amazonensis increased LHR1 transcript levels , promoted uptake of the fluorescent heme analog ZnMP , and increased the total intracellular heme content of promastigotes . Conversely , deletion of one LHR1 allele reduced ZnMP uptake and the intracellular heme pool by approximately 50% , indicating that LHR1 is a major heme importer in L . amazonensis . Viable parasites with correct replacement of both LHR1 alleles could not be obtained despite extensive attempts , suggesting that this gene is essential for the survival of promastigotes . Notably , LHR1 expression allowed Saccharomyces cerevisiae to import heme from the environment , and rescued growth of a strain deficient in heme biosynthesis . Syntenic genes with high sequence identity to LHR1 are present in the genomes of several species of Leishmania and also Trypanosoma cruzi and Trypanosoma brucei , indicating that therapeutic agents targeting this transporter could be effective against a broad group of trypanosomatid parasites that cause serious human disease .
Leishmania spp . are protozoan parasites from the Trypanosomatidae family . In mammalian hosts Leishmania is an obligate intracellular parasite , replicating as amastigotes inside acidic phagolysosomes of macrophages . Disease caused by infection with Leishmania spp . has a severe impact on human populations throughout much of the tropics . The clinical manifestations range from self-healing cutaneous lesions to lethal visceralizing disease . In many regions of the world treatment of leishmaniasis still relies on toxic drugs such as pentavalent antimony , which requires high doses and a lengthy course of treatment [1] , [2] . Treatment failure is commonly observed with pentavalent antimony [3] , and alternative drugs are costly and not widely available in endemic areas . This situation , combined with the recent increase in Leishmania infections in urban areas [4] , [5] , [6] , highlights the urgent need for identification of essential parasite molecular pathways that can be targeted by new drugs of lower cost and toxicity . Leishmania species are uniquely dependent on the acquisition of heme from the environment . Heme is a metalloporphyrin that serves as a prosthetic group for proteins that perform critical cellular functions such as oxidative metabolism , oxygen storage and transport , and signal transduction [7] . Unlike mammalian hosts which can synthesize heme [8] , Leishmania and other trypanosomatid protozoa lack several enzymes in the heme biosynthetic pathway [9] , [10] and thus depend on an exogenous supply for survival . Leishmania amazonensis acquire exogenous he3me as extracellular promastigotes and also as intracellular amastigotes replicating within macrophages [11] . The existence of a specific transporter or receptor for heme on the Leishmania plasma membrane has been speculated , based on reports showing high affinity heme binding to the cell surface of L . amazonensis promastigotes [12] and L . infantum axenic amastigotes [13] , and specific uptake of the porphyrin heme analog MgPPIX in L . donovani [14] . However , the nature of the membrane-associated molecule ( s ) responsible for heme uptake by Leishmania has remained unknown . In this study , we identify Leishmania Heme Response-1 ( LHR1 ) , a L . amazonensis gene that shares homology with HRG-4 , a C . elegans gene that encodes a plasma membrane heme importer [15] . We show that LHR1 transcript levels increase during heme deprivation , and that the LHR1 protein localizes to the plasma membrane and endocytic compartments , promotes heme uptake , and regulates the intracellular pool of heme in the parasites . Our results identify LHR1 as a strong candidate for the elusive transmembrane transporter responsible for heme acquisition from the environment by Leishmania .
The presence of hemoproteins within Leishmania amazonensis in the absence of a functional heme biosynthetic pathway [11] suggested the existence of a membrane protein capable of importing heme from the medium . As a strategy to identify this molecule , we searched the TriTryp database ( http://tritrypdb . org/tritrypdb/ ) for genes encoding transmembrane proteins with similarity to CeHRG-4 , the prototypical heme transporter from another heme auxotroph , the nematode C . elegans [15] ( WormBase Gene ID WBGene00009493 ) . In addition to BLAST homology searches , we refined our approach by identifying predicted proteins similar to HRG-4 in size and in the number of putative transmembrane domains . This search strategy identified a single open reading frame of 157 amino acids in chromosome 24 , LmjF . 24 . 2230 ( L . major ) , LmxM . 24 . 2230 ( L . mexicana ) and LinJ . 24 . 2320 ( L . infantum ) . This gene , named Leishmania Heme Response-1 ( LHR1 ) ( Genbank accession number CBZ27556 ) , shares ≈15% identity and ≈45% similarity with C . elegans HRG-4 [15] ( Figure 1A ) . We amplified a 525-bp fragment from the L . amazonensis genome using nucleotide sequences from the TriTryp database , and amino acid sequence analysis confirmed the presence of the four predicted transmembrane domains also present in CeHRG-4 . The predicted transmembrane topology suggests that the N- and C- termini of LHR1 are cytoplasmic , consistent with the proposed topology for CeHRG-4 [15] , with extracellular exposure of the conserved histidine shown to be involved in heme uptake [16] ( Figure 1B ) . Given the relatively low sequence homology between Leishmania LHR1 and CeHRG-4 , we first investigated whether LHR1 expression was influenced by heme availability . LHR1 transcript levels were elevated four fold within 15 h when L . amazonensis promastigotes were cultured in heme deficient media , compared to heme replete conditions ( Figure 2A ) . This result provided the first indication that LHR1 might be involved in heme homeostasis in Leishmania . Given the increase in LHR1 transcripts seen after cultivation in heme-deficient medium ( Figure 2A ) , we investigated whether depriving L . amazonensis promastigotes of heme for 15 h led to a subsequent increase in the uptake of fluorescent zinc mesoporphyrin IX ( ZnMP ) , a validated heme analog [17] , [18] , [19] . Low levels of ZnMP uptake were observed when the parasites were cultured in heme-containing medium , consistent with the low LHR1 transcript levels observed under these conditions . In contrast , the intracellular ZnMP fluorescence signal increased significantly after promastigotes were pre-incubated for 15 h in heme-deficient medium ( Figure 2B ) . When maintained in regular heme-containing medium and then assayed for ZnMP in the presence or absence of heme no intracellular signal was detected , showing that absence of heme during the 3–6 h period of the assay is not sufficient to promote ZnMP uptake ( not shown ) . This result is consistent with our observations , which indicate that at least 12–15 h of heme deprivation is required to upregulate LHR1 . The promastigotes remained fully viable after incubation in the absence of heme , as indicated by the viability indicator fluorescein diacetate ( FDA ) [20] ( Figure 2B ) . Thus , culture conditions that upregulate LHR1 expression lead to a concomitant increase in heme uptake . To directly examine the role of LHR1 in heme uptake by L . amazonensis , promastigote forms were transfected with an episomal expression plasmid carrying LHR1 tagged with 3xFLAG at the carboxyl terminus . Immunoblot analysis using monoclonal antibodies to the FLAG epitope detected two bands , one migrating at approximately 20 kDa corresponding to the predicted molecular mass of LHR1 , and another band at >30 kDa ( Figure 2C ) that is likely to correspond to oligomers , as previously observed with C . elegans HRG-1 [15] . Uptake of the heme analog ZnMP by promastigotes transfected with LHR1-3xFLAG was measured by flow cytometry . Compared to untransfected parasites , LHR1-transfected promastigotes showed an enhanced fluorescence signal , reflecting an increased ZnMP uptake by the parasites . These values were further increased after pre-incubation of the parasites in heme-deficient medium for 15 h to upregulate LHR1 expression ( Figure 2D ) . Importantly , LHR1 episomal expression also increased the total intracellular heme content in L . amazonensis promastigotes . The increased intracellular heme pool induced by LHR1 expression was observed in several independent experiments , performed with different numbers of promastigotes expressing LHR1-3xFLAG ( Figure 3A , B ) , or GFP-LHR1 ( Figure 4A , B ) . To determine the sub-cellular localization of LHR1 , L . amazonensis promastigotes transfected with GFP-LHR1 were cultured overnight in heme-deficient medium and then examined by confocal laser fluorescence microscopy . GFP-LHR1 was detected on the plasma membrane and in acidic intracellular compartments of promastigotes , as indicated by co-localization with lysotraker ( Figure 4C ) . We also examined mouse macrophages infected for 24 h with axenic amastigotes expressing GFP-LHR1 . The fluorescent chimeric protein was also localized at the plasma membrane of intracellular amastigotes , and in a large intracellular compartment that is likely to correspond to the megasome , the markedly enlarged lysosomal organelle typical of L . amazonensis amastigotes [21] . Parasites expressing GFP alone showed only diffuse cytosolic fluorescence , and no co-localization with lysotracker ( Figure 4C , 5A ) . After ectopic expression in HeLa cells , GFP-LHR1 localized to the plasma membrane and lysosomes , which were identified by co-localization with fluorescent dextran chased for several hours into lysosomes ( Figure 5B ) . Thus , in both L . amazonensis and mammalian cells , LHR1 is targeted to the plasma membrane and to late endosomes/lysosomes , two cellular sites from where heme can be transported into the cytosol . To obtain direct evidence for the ability of LHR1 to transport heme across membranes and make it available for essential metabolic reactions in the cytosol , we performed heterologous expression in S . cerevisiae . This unicellular eukaryote utilizes exogenous heme very poorly , even when it lacks a heme biosynthetic pathway [22] [16] . The S . cerevisiae hem1Δ strain lacks δ-aminolevulinic acid synthase ( ALAS ) , the first enzyme in the heme biosynthesis pathway . To grow , this strain requires supplementation of either δ-aminolevulinic acid ( ALA ) , the product of ALAS , or excess hemin ( ≥10 µM ) in the growth medium [23] . We found that hem1Δ requires 40-fold less hemin in the growth medium when transformed with either LHR1 or the C . elegans heme transporter CeHRG-4 ( Figure 6A ) . We also used yeast assays to measure changes in regulatory intracellular pools of heme promoted by LHR1 . We found that hem1Δ cells expressing a CYC1::lacZ promoter-reporter fusion transformed with either LHR1 or CeHRG-4 showed more than 30-fold increase in β-galactosidase activity activity ( Figure 6B ) . This is a direct indication of heme transport , since in this system CYC1 and lacZ expression depend on Hap1p , a transcription factor that binds heme [24] . Immunofluorescence microscopy showed that LHR1 was also expressed on the plasma membrane of S . cerevisiae , consistent with its subcellular localization in promastigotes and HeLa cells ( Figure 6C ) . These results show that plasma membrane-localized LHR1 is capable of conferring to hem1Δ S . cerevisiae the ability to import heme from the environment . This was directly demonstrated in kinetic experiments , which showed that LHR1 expression in wild type S . cerevisiae promotes incorporation of 55Fe-heme from the medium as a function of time ( Figure 7 ) . To genetically examine the function of LHR1 in heme transport , we generated a LHR1 mutant in L . amazonensis using homologous recombination . The linearized HYG gene replacement construct was transfected into promastigotes , and genomic DNA from hygromycin B-resistant and wild type promastigote clones was isolated , digested with XhoI and analyzed by Southern blotting . Hybridization with the LHR1 probe detected a single DNA fragment of 6013 bp in genomic DNA from wild type and two independent hygromycin-resistant clones , as expected based on the presence of restriction sites for XhoI in the upstream and downstream genes , but not in LHR1 and HYG coding sequences ( Figure 8A ) . In contrast , hybridization with the HYG probe detected a single band of 6 , 511 bp , consistent with the ≈6 , 500 bp band expected to be generated by substitution of LHR1 for HYG . These results demonstrated integration of HYG marker into the LHR1 locus , and replacement of a single LHR1 allele ( Figure 8A ) . The single knockout ( LHR1/Δlhr1 ) mutant strain showed no obvious growth defect when cultivated in regular , hemin-containing medium , with or without transfection with LHR1-3xFLAG ( Figure 8B ) . However , following incubation in heme-depleted medium , a condition that upregulates LHR1 expression ( Figure 2A ) , the level of ZnMP uptake over 2 h was reduced by ≈50% in LHR1/Δlhr1 when compared to wild type ( Figure 8C ) . The total intracellular heme content was also reduced by ≈50% in LHR1/Δlhr1 promastigotes , a phenotype that was partially restored by transfection of LHR1/Δlhr1 parasites with the LHR1-3xFLAG episomal plasmid ( Figure 8D , E ) . Repeated attempts were made to replace the second LHR1 allele with a NEO deletion construct without recovery of viable clones , suggesting that LHR1 is an essential gene in L . amazonensis . When attempts were made to delete the second LHR1 allele in LHR1/Δlhr1 parasites transfected with the LHR1-3xFLAG plasmid , the only viable promastigotes triple resistant to hygromycin , neomycin and nourseothricin ( resistance conferred by the episomal LHR1-3xFLAG plasmid ) that were recovered still had one endogenous LHR1 copy ( data not shown ) . These results suggest that the levels of heme transport conferred by episomal expression of LHR1-3xFLAG are not sufficient to sustain promastigote growth , even when an abundant source of heme is provided in the culture medium . This observation is consistent with the incomplete rescue of the heme uptake and homeostasis phenotypes of LHR1/Δlhr1 L . amazonensis by LHR1-3xFLAG expression ( Figure 8 ) . Taken together , these results strongly suggest that LHR1 accounts for the majority of the heme transport activity of L . amazonensis .
The existence of an essential pathway for acquisition of exogenous heme in Leishmania and other trypanosomatid protozoa was proposed decades ago [9] , when it became clear that these organisms lack several enzymes of the heme biosynthetic pathway [10] . However , the molecule ( s ) responsible for this critical activity remained unknown . In this work we identify LHR1 , a Leishmania gene upregulated under heme-deficient conditions that encodes a membrane protein able to promote heme uptake from the environment . Transfection of Leishmania with LHR1 promotes uptake of a heme analog and increases the total intracellular heme concentration in the parasites . A L . amazonensis single-allele LHR1 knockout strain shows reduced uptake of a heme analog and has a significantly smaller intracellular heme pool . Viable parasites lacking both chromosomal copies of LHR1 could not be isolated even when carrying episomal LHR1 , suggesting that LHR1 performs a critical function that depends on levels of expression not achieved with the tagged LHR1-3xFLAG . Importantly , LHR1 functionally complemented a yeast strain deficient in heme biosynthesis , in both growth and heme-dependent gene expression assays . These results strongly support a role of LHR1 as a transporter and not a receptor for heme , because yeast cells lack a efficient heme import machinery [22] [16] . The efficiency of heme uptake from the environment may vary among Leishmania species , since Campos-Salinas et al . reported faster incorporation of Mg-PPIX by L . donovani promastigotes [14] than what we observed with L . amazonensis . Future studies may provide evidence for the intriguing possibility that LHR1 is differentially expressed in visceral strains of Leishmania , a property that might be associated with their increased virulence and capacity to proliferate in internal organs . LHR1 was identified based on its partial sequence identity and similarity to HRG-4 , a gene encoding a plasma membrane heme transporter in the nematode C . elegans . HRG-4 was identified in a genetic screen designed to take advantage of the heme auxotrophy of C . elegans to identify heme-responsive genes [15] . LHR1 and CeHRG-4 [15] [16] have a similar molecular mass ( ∼20 kDa ) , and four predicted transmembrane domains . One intriguing difference observed between CeHRG-4 and LHR1 is their subcellular localization . While HRG-4 is localized primarily on the plasma membrane , GFP-tagged LHR1 was detected on the plasma membrane and on endocytic compartments of L . amazonensis . In mammalian cells , GFP-LHR1 was also targeted to the plasma membrane and lysosomes , strongly suggesting that the large intracellular compartments containing LHR1 in L . amazonensis correspond to parasite lysosomal compartments . A morphometric and cytochemical study in L . amazonensis showed that during differentiation of promastigotes into amastigotes , the lysosomal vacuoles of promastigotes become a megasome , a very large compartments that can comprise up to ∼5% of the total cell volume [21] . This stage-specific lysosomal pattern is very consistent with the intracellular localization of GFP-LHR1 in our study . In addition to the plasma membrane , GFP-LHR1 was observed in several intracellular vesicles in promastigotes and in one very large compartment in intracellular amastigotes . The dual localization of LHR1 on the plasma membrane and on lysosomes raises interesting questions about the cellular site where heme is translocated into the cytosol . In yeast , LHR1 was targeted to the plasma membrane and promoted functional complementation of a strain incapable of synthesizing heme . However , earlier studies in Leishmania donovani showed that hemoglobin is internalized and degraded within parasite lysosomes [25] , releasing heme that can then be transported into the cytosol to promote parasite growth . Interestingly , exogenously added hemin rescued the growth of a L . donovani strain defective in endocytic transport into lysosomes , indicating that heme translocation across the membrane can occur at both locations – the plasma membrane and the parasite lysosome [26] . The ATP-binding cassette protein LABCG5 was also recently proposed to mediate the salvage of heme released after lysosomal degradation of internalized hemoglobin in L . donovani . This intracellular process for heme salvage from degraded hemoglobin was proposed to be distinct from the pathway directly promoting porphyrin transport into the parasites [14] . Additional studies should determine if LHR1 can also mediate the uptake of heme released from hemoglobin inside parasite lysosomes , or if it's primary role is to transport heme from the environment directly across the plasma membrane . LHR1 null strains could not be generated despite extensive attempts , suggesting that this transporter is essential for the survival of promastigote forms of L . amazonensis . Episomal expression of LHR1 increased the intracellular heme concentration of wild type and single knockout L . amazonensis promastigotes , but was not sufficient to allow recovery of viable parasites lacking both copies of the gene . This finding is likely related to the fact that episomal LHR1 expression failed to restore the intracellular heme concentration to the same levels observed in wild type parasites . Dysregulated expression and incomplete functional complementation is a frequent observation after episomal or integrated gene expression in Leishmania [27] , [28] , [29] , [30] . Incomplete restoration of heme acquisition in LHR1 double knockout parasites may result in the impairment of critical , essential roles played by hemoproteins in the parasites . For example , LFR1 , the recently identified NADPH-dependent ferric iron reductase from L . amazonensis , contains a bis-heme motif that is essential for activity and required to allow iron acquisition through the ferrous iron transporter LIT1 [30] . Thus , deleting both copies of LHR1 may severely affect not only heme uptake , but also the ferrous iron acquisition process . Searches of the TriTryp database indicate that highly syntenic , close homologs of L . amazonensis LHR1 ( LmxM . 24 . 2230 ) are present in the additional Leishmania species L . major ( LmjF24 . 2230 ) , L . braziliensis ( LbrM . 24 . 2310 ) and L . infantum ( LinJ . 24 . 2320 ) , and in the Trypanosoma species T . brucei ( Tb427 . 08 . 6010 , Tb927 . 8 . 6010 ) , T . brucei gambiense ( Tbg972 . 8 . 6030 ) , T . congolense ( TclL3000 . 8 . 5780 ) , and T . cruzi ( Tc00 . 1047053511071 . 190 ) . These trypanosomatid species are the causative agents of serious infectious diseases in humans ( L . infantum , visceral leishmaniasis; T . brucei gambiense , sleeping sickness; T . cruzi , Chagas' disease ) or in livestock ( T . congolense and T . brucei brucei , cattle Nagana ) . Given that the human genome does not include putative orthologs of CeHRG-4 and LHR1 [15] , our study suggests that LHR1 may represent a promising target for the development of new therapeutic drugs with a potentially broad impact in human health and quality of life .
The L . amazonensis IFLA/BR/67/PH8 strain was provided by Dr . David Sacks ( Laboratory of Parasitic Diseases , NIAID , NIH ) . Promastigotes were cultured at 26°C in promastigote growth medium: M199 ( Gibco BRL ) pH 7 . 4 supplemented with 10% heat-inactivated FBS , 5% penicillin-streptomycin , 0 . 1% hemin ( 25 mg/ml in 50% triethanolamine ) , 10 mM adenine ( pH 7 . 5 ) and 5 mM L-Glutamine , as previously described [30] . Heme-depleted FBS was generated by treating heat inactivated FBS with 10 mM ascorbic acid for 4 h , followed by dialysis in PBS overnight and filter-sterilization . Heme depletion was verified by measuring the optical absorbance at 405 nm [31] . Parasite viability was assessed by fluorescent microscopy using fluorescein diacetate ( FDA; Invitrogen ) in combination with Propidium Iodide ( PI; Sigma-Aldrich ) , as described in [20] . A single open reading frame , LmjF . 24 . 2230 ( L . major ) , LmxM . 24 . 2230 ( L . mexicana ) , or LinJ . 24 . 2320 ( L . infantum ) was identified through BLAST homology searches of the Leishmania database ( TritrypDB ) . Forward ( 5′GGATCCATGAACGAGTTGGAGCGC ) and reverse ( 5′GGATCCCTATGCACAGTTCTCC-3′ ) primers ( added BamHI sites underlined ) were used to amplify the 525 base pair ORF from genomic DNA of L . amazonensis . The PCR product was cloned into the pCR2 . 1-TOPO ( Invitrogen ) to generate a pCR-LHR1 plasmid , and the coding sequence was confirmed by sequencing . To construct a GFP-LHR1 gene fusion , pCR2 . 1-LHR1 digested with BamH1 was cloned into pXGGFP2+ ( courtesy of Dr . S . Beverley , Washington University ) , which drives expression in Leishmania of proteins fused to GFP at the N-terminus [32] to yield the pXG-GFP+LHR1 plasmid . After transfection , clones resistant to G418 ( 50 µg/ml ) were isolated . To generate LHR1 tagged with 3xFLAG at the carboxyl-terminus , the primers 5′GGATCCACCATGAACGAGCGCAAGCG ( forward ) and 5′GGATCCCTATCGCGATGCACAGTTCTCCTTTGAC ( reverse ) ( BamHI sites underlined , NruI site in italics ) were used to amplify the LHR1 ORF from pCR2 . 1-LHR1 . The modified LHR1 ORF containing a NruI site before the stop codon and flanked by BamHI restriction sites was cloned into pCR2 . 1 using TOPO PCR Cloning kit ( Invitrogen ) to generate plasmid pCR2 . 1-LHR1 ( Nrul-stop ) . A NruI excised fragment of the 3xFLAG epitope tag ( Sigma-Aldrich ) was cloned into NruI digested pCR2 . 1-LHR1 ( Nrul-stop ) to generate pCR2 . 1-LHR1-3xFLAG . pCR2 . 1-LHR1-3xFLAG was digested with BamHI and cloned into pXG-SAT ( courtesy of Dr . S . Beverley , Washington University ) to yield the pXGSAT-LHR1-3xFLAG plasmid . After transfection and isolation of clones resistant to 50 µg/ml nourseothricin , LHR1-3xFLAG expression was detected by immunoblot [30] . Total protein extracts ( 50 µg ) from wild type promastigotes expressing either LHR1 or LHR1-3xFLAG were separated on 15% SDS-PAGE , transferred to Immobilon membrane , blocked in 3% nonfat dry milk in TBS ( 50 mM Tris , 0 . 138 M NaCl , 2 . 7 mM KCl , pH8 ) followed by detection with the mouse monoclonal antibody M2 that recognizes the FLAG epitope ( Sigma-Aldrich ) . The mammalian expression plasmid GFP-LHR1 was generated by amplifying GFP-LHR1 from pXG-GFP2+-LHR1 plasmid described above , with primers annealing to the start codon of GFP ( forward ) and to the stop codon of LHR1 ( reverse ) . The resulting GFP-LHR1 fragment was cloned into pCR2 . 1-TOPO to create pCR2 . 1-KpnI-GFP-LHR1-XhoI , double digested with KpnI XhoI , and the GFP-LHR1 fragment cloned into the KpnI and XhoI sites of pShuttle-CMV [32] to yield pShuttleCMV-GFP-LHR1 . Gene deletion constructs containing hygromycin B phosphotransferase ( HYG ) or neomycin phosphotransferase ( NEO ) were based on the expression vectors pXG-HYG and pXG-NEO ( courtesy of Dr . S . Beverley , Washington University , St . Louis , MO ) . A 1 , 000-bp flanking sequence upstream of L . amazonensis LHR1 was amplified using primers: 5′-GTTGGGCGACTTGTACGG-3′ UPLHR1 ( forward ) and 5′-GGATCCCGGGTCAACCAAATGCGGAAC-3′ UPREVLHR1 ( reverse ) . A 2 , 500-bp flanking sequence downstream of LHR1 was amplified using primers: 5′-GGATCCCGGGCTTGGCCTCATTGATTC-3′ DOLHR1 ( forward ) and 5′-CCTGTGAAGATGTTCC-3′ DOREVLHR1 ( reverse ) . The upstream and downstream sequences were cloned into pCR2 . 1-TOPO vector ( Invitrogen ) creating the plasmid pCR2 . 1-UP-LHR1 and pCR2 . 1-DOWN-LHR1 . The upstream region was excised from pCR2 . 1-UP-LHR1 plasmid with BamHI and cloned into pCR2 . 1-DOWN-LHR1 linearized with BamHI . The resulting plasmid pCR2 . 1-UP-LHR1-DOWN contained two SmaI sites at the junction of the upstream and downstream sequences . To generate the deletion constructs , the HYG and NEO ORFs were amplified using primers: 5′-GTTCCGCATTTGGTTGGATGAAAAAGCCTGAAC-3′ HYG5primeUTR of LHR1 ( HYG forward ) 5′-TCAATGAGGCCAAGCCCTATTCCTTTGCCCT-3′ HYG3primeUTR of LHR1 ( HYG reverse ) and 5′-GTTCCGCATTTGGTTGGATGGGATCGGCCATTG-3′ NEO5primeUTR of LHR1 ( NEO forward ) 5′-TCAATGAGGCCAAGCCTCAGAAGAACTCGTCAA-3′ NEO3primeUTR of LHR1 ( NEO reverse ) ( underlined sequences corresponding to the SmaI site junctions of pCR-UP-LHR1-DOWN ) from the pXG-based vectors . Fusion of the ends of amplified HYG and NEO ORFs to the homologous ends of pCR2 . 1-UP-LHR1-DOWN linearized with SmaI was carried out using Clontech In-Fusion PCR Cloning according to the manufacturer's instructions . DNA fragments from the gene deletion constructs were released from the vector backbone with BamHI and XhoI and gel-purified . Transfections of L . amazonensis promastigotes were performed as previously described [30] . To create a Δlhr1 single knockout strain , the region containing the LHR1 gene was replaced with the hygromycin B phosphotransferase gene ( HYG ) , and clones resistant to hygromycin B ( 100 µg/ml ) were isolated . Southern blots were performed to determine integration of the HYG marker at the LHR1 locus . Genomic DNA was isolated in TELT lysis buffer as described [33] , digested with XhoI overnight , separated on 0 . 8% agarose and transferred to a nylon membrane . Blots were blocked and hybridized to digoxigenin ( DIG ) labeled PCR probes according to the manufacturer's protocol ( Roche ) . The DIG labeled probes for HYG and LHR1 were generated by PCR amplification with primers: HYG: 5′- ATGAAAAAGCCTGAAC-3′ ( forward ) and 5′- CTATTCCTTTGCCCT-3′ ( reverse ) ; LHR1: 5′GGATCCACCATGAACGAGCGCAAGCG ( forward ) and 5′ GGATCCCTATCGCGATGCACAGTTCTCCTTTGAC ( reverse ) , using the manufacturer's protocol ( Roche ) . The LHR1/Δlhr1 single knockout strain was cultured in promastigote growth medium supplemented with 100 µg/ml hygromycin B . Transfection of LHR1/Δlhr1 with the LHR1-3xFLAG episomal expression plasmid was carried out as described above . The LHR1 gene deletion procedure was repeated using the NEO construct with or without prior transfection with LHR1-3xFLAG . Transfected L . amazonensis promastigote clones expressing GFP-LHR1 were selected in 50 µg/ml G418 . To visualize LHR1 , log phase promastigotes expressing GFP-LHR1 were imaged live on an UltraView Vox spinning disk confocal system ( Perkin Elmer ) equipped with an electron multiplier CCD camera ( C9100-50; Hamamatsu ) . Images were acquired and processed using the Volocity software suite ( PerkinElmer ) . For colocalization experiments , parasites were incubated with 1 µM Lysotracker red ( InVitrogen ) for 20 min at room temperature in serum-free M199 and washed twice before imaging . For LHR1 localization in the intracellular amastigote forms , bone marrow macrophages ( prepared from C57BL/6 mice , as described in ( 25 ) ) infected with L . amazonensis axenic amastigotes ( differentiated from promastigotes transfected with the GFP vector , GFP-LHR1 ) were fixed with 2% PFA and treated with 0 . 1 mg/ml RNase A for 1 h . Samples were washed three times with PBS , stained for 1 min with 50 µg/ml PI , followed by three washes with PBS . Coverslips were mounted in ProLong ( Molecular Probes , Invitrogen ) and imaged on a confocal microscope ( Leica TCS SP5 X ) using the application suite software ( Leica ) , followed by image processing with the Volocity software suite . HeLa CCL-2 cells ( HeLa 229 ) were seeded on 35-mm MatTek dishes ( 2 . 0 ml of 0 . 75×105 cells/ml ) 24 h before experiments in DMEM 10% FBS and 1% penicillin/streptomycin , and incubated at 37°C and 5% CO2 . The media was replaced with 250 µl Opti-MEM I reduced serum and cells were transfected with Lipofectamine and 1 mg pShuttleCMV-GFP-LHR1 , according to the manufacturer's instructions ( Invitrogen ) . To label lysosomes , 0 . 5 mg/ml Texas Red dextran ( MW 10 , 000 , Sigma-Aldrich ) was added to cells followed by incubation at 37°C for 1 h , several PBS washes and a 2 h chase at 37°C , as previously described [34] . For live imaging , dishes were placed in an environmental chamber ( LiveCell System; Pathology Devices , Inc . ) at 37°C with 5% CO2 attached to an UltraView VoX spinning-disk confocal system ( PerkinElmer ) equipped with a CCD camera ( C9100-50; Hamamatsu ) and processed using Volocity software suite ( PerkinElmer ) . For imaging in S . cerevisiae , yeast transformants were cultivated to mid-log phase in 2% w/v raffinose SC ( -Ura ) medium supplemented with 0 . 4% w/v galactose and 250 µM ALA , and fixed with 4% formaldehyde for 1 h at room temperature . Immunofluorescence microscopy was performed as described elsewhere [35] , and images were acquired using a DMIRE2 epifluorescence microscope ( Leica ) connected to a Retiga 1300 cooled Mono 12-bit camera . Log phase growth wild-type L . amazonensis promastigotes were washed twice with PBS , resuspended at 106 parasites/ml in promastigote medium without hemin and 20% heme-depleted FBS , or regular promastigote medium with hemin and 20% untreated FBS . After 15 h at 26°C , a total of 108 promastigotes were collected . Three independent samples were used to isolate total RNA using Qiagen RNAeasy kit ( Qiagen ) according to the manufacturer's instructions . cDNA synthesis was carried out using 1 µg of total RNA and Superscript II Reverse Transcriptase ( Invitrogen ) according to the manufacturer's protocol . To quantify LHR1 transcript levels in each sample , 1 µl of the synthesized cDNA was amplified using LHR1 specific primers: sense 5′-CGCTCGTACTTTTGTGGA-3′ and antisense 5′-CCTGAATCAATGAGGCCAAG-3′ , and GAPDH specific primers: sense 5′-GAAGTACACGACCTTCTTC and antisense 5′-CGCTGATCACGACCTTCTTC as the reference gene . Quantitative real time PCR was performed using a BioRad iCyler iQ Real-Time PCR System ( BioRad Laboratories ) using the SYBR green fluorophore , according to the manufacturer's instructions . All reactions were performed in triplicate , and no template DNA was included in each experiment as a negative control . The cycle threshold ( Ct ) value was determined , and the fold induction of LHR1 transcripts was calculated using the 2−ΔΔCt method [36] . Heme ( iron protoporphyrin IX ) concentration was determined using the pyridine hemochrome method [37] . Log phase growth promastigotes cultivated in regular promastigote growth medium were collected by centrifugation , counted , washed once with PBS , resuspended in 1 ml of 1 mM Tris-HCl pH 8 . 0 , and sonicated for 2 min in an ice water bath using a Branson digital sonifier at 30% power setting , in pulses of 5 seconds intercalated with 5 s of cooling . Aliquots of 840 µl were transferred to 13×100 mm glass tubes , 100 µl of 1 N NaOH was added followed by vortexing , and after 2 min 200 µl of pyridine solution ( Sigma-Aldrich ) was added to each sample followed by vortexing . Samples were then transferred to a 1 ml cuvette and a baseline absorbance spectrum was obtained . A few sodium dithionite crystals ( 2–3 mg ) were added to the sample , and the reduced hemochrome absorbance spectrum between 500 and 600 nm was acquired after 1 min . Heme concentrations were calculated based on the millimolar extinction coefficient of 20 . 7 , for the difference in absorption between the spectrum peak at 557 nm and the valley at 541 nm . Zn ( II ) Mesoporphyrin IX ( ZnMP ) ( Frontier Scientific ) was dissolved in DMSO at 10 mM , and the uptake assay was performed as described [14] , with modifications . Wild type and LHR1-3xFLAG-expressing L . amazonensis promastigotes ( 107 parasites per ml ) were cultured in regular or heme-deficient medium ( M199 , 10% heme-depleted FBS , 1% Pen/Step ) for 15 h . 2×109 parasites were collected by centrifugation , washed once with PBS and divided equally into regular promastigote growth medium or heme-deficient medium containing 10 µM ZnMP for 3 or 6 h . The ZnMP uptake reaction was stopped by adding an equal volume of ice-cold 5% BSA in PBS , parasites was collected by centrifugation , washed , and the fluorescence intensity of ZnMP accumulation in a total of 3×104 parasites was measured by flow cytometry ( BD FACSCanto , excitation at 488 nm and emission equal or greater than 670 nm ) [15] , or live-imaged in a Nikon E200 equipped with a DS-Fi1 camera and Digital Sight software . Flow cytometry data were analysed using FlowJo 6 . 3 software ( Tree Star , Inc . ) . | The biological activity of many proteins and enzymes requires heme , a large organic ring containing one iron atom at the center . It has been known for several decades that trypanosomatid protozoa lack several enzymes in the heme biosynthetic pathway . Therefore , unlike mammalian cells that can synthesize heme , these unicellular organisms must acquire heme from the environment . However , the mechanism by which this critical co-factor is transported into trypanosomatid parasites was unknown . In this study we identified LHR1 , a trans-membrane protein from Leishmania amazonensis that mediates transport of extracellular heme into the parasites . Parasites partially deficient in LHR1 are impaired in heme import , and strains completely deficient do not survive . Genes highly similar to LHR1 are present in several species of trypanosomatid parasites that cause human disease , identifying this transporter as an important target for the development of anti-parasitic drugs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biology"
] | 2012 | Heme Uptake by Leishmania amazonensis Is Mediated by the Transmembrane Protein LHR1 |
The macrophage is the primary host cell for the fungal pathogen Histoplasma capsulatum during mammalian infections , yet little is known about fungal genes required for intracellular replication in the host . Since the ability to scavenge iron from the host is important for the virulence of most pathogens , we investigated the role of iron acquisition in H . capsulatum pathogenesis . H . capsulatum acquires iron through the action of ferric reductases and the production of siderophores , but the genes responsible for these activities and their role in virulence have not been determined . We identified a discrete set of co-regulated genes whose transcription is induced under low iron conditions . These genes all appeared to be involved in the synthesis , secretion , and utilization of siderophores . Surprisingly , the majority of these transcriptionally co-regulated genes were found clustered adjacent to each other in the genome of the three sequenced strains of H . capsulatum , suggesting that their proximity might foster coordinate gene regulation . Additionally , we identified a consensus sequence in the promoters of all of these genes that may contribute to iron-regulated gene expression . The gene set included L-ornithine monooxygenase ( SID1 ) , the enzyme that catalyzes the first committed step in siderophore production in other fungi . Disruption of SID1 by allelic replacement resulted in poor growth under low iron conditions , as well as a loss of siderophore production . Strains deficient in SID1 showed a significant growth defect in murine bone-marrow-derived macrophages and attenuation in the mouse model of infection . These data indicated that H . capsulatum utilizes siderophores in addition to other iron acquisition mechanisms for optimal growth during infection .
Iron acquisition is critical to cellular function and survival . During infection of mammals , the host limits the access of iron to microbial pathogens by a variety of means [1] . In turn , pathogens utilize a number of strategies to acquire iron in the face of iron restriction by the host . Here we investigate the role of siderophore-mediated iron acquisition in the fungal pathogen Histoplasma capsulatum , which parasitizes host macrophages during infection . Histoplasma capsulatum is a dimorphic , fungal pathogen that causes respiratory and systemic disease in humans . Infection of mammals is initiated by inhalation of fungal spores from soil in regions of the U . S . where the organism is endemic . Once in the host , H . capsulatum grows in a budding yeast form that colonizes alveolar macrophages . Yeast cells replicate within the macrophage phagolysosome , but the molecular mechanisms governing survival within host cells remain largely undefined [2] . Replication of H . capsulatum within macrophages is dependent on the availability of iron . During infection of murine peritoneal and human monocyte-derived macrophages , addition of the iron chelator deferoxamine inhibits the intracellular growth of H . capsulatum . This inhibition is suppressed by the addition of iron-rich transferrin ( holotransferrin ) , indicating that active iron acquisition is a critical determinant of intracellular growth [3] , [4] . In addition , the adaptive immune response to H . capsulatum infection , which triggers production of the cytokine interferon-gamma ( IFNγ ) by T-cells [5] , may curtail intracellular fungal growth by limiting iron acquisition by the fungus . Treatment of murine peritoneal macrophages with interferon-gamma ( IFNg ) causes growth restriction of H . capsulatum which , in turn , can be reversed by addition of holotransferrin [3] . IFNγ downregulates surface transferrin receptors , suggesting that a major means by which IFNγ inhibits intracellular fungal growth is via iron limitation . These studies indicate that iron acquisition plays a critical role in Histoplasma virulence . However , although several biochemical activities have been identified in H . capsulatum , little is known about the genes that regulate iron accumulation . H . capsulatum is known to secrete hydroxamate siderophores that act as low-molecular-weight ferric iron chelators under low iron conditions [6] , [7] . Additionally , iron limitation also induces reductive iron assimilation , including ferric reductase activity [8] . To assess the role of these iron acquisition mechanisms in virulence , we identified H . capsulatum genes that function in siderophore-mediated iron acquisition . These genes were transcriptionally induced under low iron conditions and contained a common putative regulatory site in their upstream regions that might govern their coordinate expression . Inspection of the sequences revealed that these genes were clustered together in the genome , and thus defined a secondary metabolite gene cluster involved in siderophore biosynthesis . Disruption of one of the genes of this pathway , SID1 , resulted in elimination of hydroxamate siderophore synthesis , diminished growth of H . capsulatum within macrophages , and compromised virulence in mice . These data indicated that iron scavenging though siderophore production facilitates the parasitic growth of H . capsulatum .
The genes involved in iron acquisition are tightly regulated at the level of transcription in most microorganisms . To identify genes that are transcriptionally regulated by iron limitation conditions in H . capsulatum , we grew H . capsulatum ( strain G217B ) yeast cells under iron-limiting ( 100 µM deferoxamine mesylate ) or iron-replete ( 5 or 10 µM FeSO4 ) conditions for various periods of time . The transcriptional profiles of these samples were compared using whole-genome oligonucleotide microarrays . We found seven genes that were transcriptionally induced under conditions of iron limitation and named them according to the putative functions of the corresponding proteins ( Figure 1A ) . These genes included L-ornithine monooxygenase ( SID1 , previously named LOM1 [9] , EU253976 ) , the enzyme catalyzing the first committed step in extracellular and intracellular siderophore production in other organisms; an acetylase ( SID3 , EU253977 ) ; an acid co-A ligase ( SID4 , EU253978 ) ; a non-ribosomal peptide synthase ( NPS1 , EU253973 ) ; an oxidoreductase ( OXR1 , EU253974 ) ; a major facilitator superfamily ( MFS ) transporter ( MFS1 , EU253970 ) [10]; and an ATP-binding cassette transporter ( ABC1 , EU253969 ) [10] . Based on precedent from other fungi , these genes were likely to have roles in the production , transport , and utilization of siderophores , and could comprise the entire pathway of hydroxamate siderophore production [11]–[13] . SID3 is orthologous to SidF in Aspergillus fumigatus , an N5-transacylase involved in the synthesis of fusarinine and triacetylfusarinine [13] . NPS1 is orthologous to the non-ribosomal peptide synthases involved in extracellular siderophore production in ascomycetes [13] , [14] . MFS1 is homologous to mirB of Aspergillus nidulans , a siderophore transporter [15] . And finally , ABC1 contains homology to MTABC3 , a mammalian ABC transporter involved in iron homeostatis [16] . Quantitative RT-PCR ( QRT-PCR ) analysis confirmed that all seven genes were transcriptionally induced during iron deprivation ( Figure 1B ) . Both the microarrays and QRT-PCR analysis also showed that transcription of these genes began to be induced even in the presence of iron supplementation at late time points , suggesting that iron eventually becomes limiting under these conditions , or perhaps that the induction of genes involved in siderophore synthesis is highly sensitive to a subtle decrease in iron availability . Surprisingly , all but one of these iron-regulated genes were located adjacent to each other in an approximately 25 kb region of the genome of the G217B strain of H . capsulatum ( Figure 2A ) . Only MFS1 is encoded by a distinct genomic region . This genomic cluster of iron-regulated genes was also present in the two other sequenced strains of H . capsulatum: G186AR ( Genome Sequencing Center , Washington University , St . Louis , MO ) and WU24 ( Broad Institute , Massachusetts Institute of Technology , Cambridge , MA ) . In G186AR , an additional MFS transporter , MFS2 ( EU253971 ) , as well as a second acetylase ( SID5 , EU253979 ) , was found within the putative iron-regulated gene cluster . WU24 contained the 5′ end of the MFS2 gene , but did not contain SID5 . We observed a similar genome structure in the closely related fungus Coccidioides immitis ( data not shown ) , but only a subset of the genes were found adjacent to each other in other fungal genomes ( see discussion ) . Two of the genes located within this genomic cluster , NIT22 ( encoding a dehydratase , [9] , EU253972 ) and RTA1 ( encoding a predicted protein containing an RTA domain , resistance to 7-aminocholesterol , EU253975 ) , did not appear to be iron regulated by microarray . We used QRT-PCR analysis to determine whether these genes were regulated differentially in response to iron levels . However , whereas RTA1 expression was not detected by QRT-PCR under either iron-rich or iron-poor conditions ( data not shown ) , NIT22 expression was induced under iron limitation ( Figure 1B ) . Unlike the other iron-regulated genes in this genomic cluster , NIT22 did not encode a protein with homology to known siderophore-biosynthesis genes . Thus , its induction under conditions of iron limitation may reflect a previously unidentified role of its dehydratase activity in siderophore biosynthesis . To determine if the iron-dependent regulation of this gene cluster extended beyond the genes we identified by microarray , we also examined the expression of UBP1 ( encoding a ubiquitin C-terminal hydrolase ) , a gene approximately 6kb upstream of OXR1 . UBP1 expression was not regulated by iron levels ( Figure 1B ) . In addition , we used QRT-PCR to determine that genes found in the iron cluster of G186AR were also transcriptionally regulated by iron levels ( Figure S1 ) . Since the genes in this siderophore biosynthetic cluster were transcriptionally co-regulated in response to iron levels , we used Multiple Em for Motif Elicitation ( MEME ) to identify conserved sequences in the upstream regions of these genes from both the G217B and G186AR strains of H . capsulatum . A consensus site , 5′- ( G/A ) ATC ( T/A ) GATAA-3′ , was present at least once in the 5′ regions of all of the genes in the siderophore biosynthetic cluster ( Figure 2B ) . Interestingly , this consensus contained an HGATAR sequence , the recognition site for fungal GATA transcriptional regulators [17] , some of which are known to regulate the expression of siderophore-biosynthesis genes in other fungi [18]–[20] . In H . capsulatum , an ortholog of these GATA factors , Sre1 , was shown to bind in vitro to the consensus site we identified here , suggesting that Sre1 limits expression of the siderophore biosynthetic gene cluster under iron-replete conditions ( Chao et al , submitted ) . To test whether this pathway was important for siderophore production , we disrupted SID1 , which encodes the enzyme that catalyzes the first committed step in siderophore production . The gene disruption was performed in the wild-type G186ARura5Δ strain background and confirmed by PCR ( data not shown ) and Southern blot analysis ( Figure 3A ) . The sid1Δ strain grew with normal kinetics under iron-replete conditions , but displayed a growth defect under conditions of iron limitation ( Figure 3B ) . Consistent with its poor growth under iron-limiting conditions , the sid1Δ strain also failed to produce siderophores ( Figure 3C ) . To insure that the observed growth defect was due to disruption of SID1 , we complemented the sid1Δ strain with a wild-type copy of the SID1 gene . SID1 , including all intergenic regions 5′ and 3′ of the open reading frame , was cloned into an integrating Agrobacterium tumefaciens T-DNA vector ( pED2 ) . This construct was randomly integrated into the wild-type genome and into two independent sid1Δ isolates . The SID1 gene fully complemented the growth defect in low-iron medium ( Figure 3B ) , but only partially complemented siderophore production ( Figure 3C ) for unknown reasons . QRT-PCR analysis revealed that SID1 transcript accumulation was similar under iron-limiting conditions in the wild-type and complemented strains ( data not shown ) . We also introduced a SID1 complementation construct on a high-copy episomal plasmid , but these transformants expressed SID1 at very low levels and showed little complementation of siderophore production ( data not shown ) , suggesting that integration into the genome may be required for normal levels of SID1 transcription . In other pathogens , iron acquisition from the host is a critical virulence determinant . To examine the role of siderophore-mediated iron acquisition in H . capsulatum infection , we infected bone-marrow-derived murine macrophages ( BMDMs ) with wild-type , sid1Δ , and complemented strains . At zero , 6 , 24 , and 48 hours following infection , the macrophages were lysed and colony-forming units ( CFUs ) of H . capsulatum cells were determined ( Figure 4A ) . Wild-type H . capsulatum cells were able to proliferate within the macrophages as expected . Although the sid1Δ strains were also able to grow intracellularly , they reached only ∼40% of wild-type levels by 48 hours after infection ( p<0 . 001 ) . The doubling time of wild-type cells in macrophages was 10 hours while the sid1Δ strain doubled in 14 hours . Addition of 100 µM FeSO4 to infected macrophages reversed this phenotype , strongly suggesting that the proliferation defect in macrophages was due to the decreased ability of the mutant strains to acquire iron ( Figure 4B ) . The complemented strains did not display a growth defect in macrophages , indicating that they made sufficient siderophores to permit wild-type growth during macrophage infection ( Figure 4A ) . The growth defect of the mutant strain and complementation of the defect with either reconstitution of the wild-type gene or the addition of iron was consistent over several infections ( Figure 4C ) . To investigate whether siderophore production plays a role in animal infections , we first infected C57BL/6J mice intranasally with a lethal dose ( effective inoculum of ∼1×105 CFU ) of either wild-type , sid1Δ , or complemented strains . No significant differences in pulmonary fungal burden were observed under these conditions ( data not shown ) . However , since the literature suggested that assessing the virulence of a mutant strain in competition with wild-type can provide a more sensitive assay for pathogenesis defects [21] , [22] , we decided to perform a competitive infection with wild-type , sid1Δ , and the complemented strains . We infected C57BL/6J mice intranasally with a sublethal dose ( effective inoculum of ∼1×104 CFU ) of an equal mix of either wild-type and sid1Δ or wild-type and sid1Δ+SID1 strains . Lung homogenates from multiple time points were analyzed for CFUs and ratios of wild-type , sid1Δ , and complemented strains were calculated . The competitive index , which reflects the defect of a particular strain with respect to the wild-type strain , was determined as described in Materials and Methods . We observed that the sid1Δ strain showed a significant defect in pulmonary colonization compared to wild-type cells ( Figure 5 ) . This defect began at day 5 , with a severe defect apparent at day 15 ( p<0 . 01 ) . This phenotype was reversed in the complemented strain , which accumulated to slightly higher levels than the wild-type strain for unknown reasons ( p<0 . 05 ) . These data indicated that siderophore production is important for optimal growth of H . capsulatum in mouse lungs .
Iron acquisition is essential for the growth of most microorganisms and occurs primarily by two mechanisms: siderophore production and reductive iron assimilation [23] . Both of these processes are induced by H . capsulatum during iron limitation [6] , [8] , but their role in cell growth and virulence has not been previously investigated . In this study , we determined the role of siderophore production in growth under conditions of iron limitation as well as during infection . We identified a siderophore biosynthetic gene cluster that was transcriptionally induced in response to iron limitation . Disruption of siderophore production resulted in iron-dependent growth in culture and during macrophage infection , and caused a growth defect in mice . Presumably the sid1Δ mutant shows a strong pulmonary colonization defect in competition with wild-type because it has a reduced capacity for iron acquisition in vivo . The in vivo growth defect caused by lack of siderophore production is most pronounced at 15 days post infection . Interestingly , the peak of IFN-γ production by T cells in response to H . capsulatum infection occurs at day 14 [24] . Since one of the functions of IFN-γ is to restrict iron , perhaps siderophore production is primarily required for iron acquisition during the latter stages of infection , whereas redundant iron acquisition mechanisms , such as reductive iron assimilation , might play a role early in infection . In other fungal pathogens , siderophore production and reductive iron assimilation , which is usually mediated by ferric reductase activity , play differential roles in pathogenesis . For example , siderophore production is essential for virulence in Aspergillus fumigatus , but reductive iron assimilation is neither necessary nor sufficient for normal growth and survival in the host [25] . Similarly , in the phytopathogens Cochliobolus miyabeanus , Alternaria brassicicola , Cochliobolus heterostrophus , and Fusarium graminearum , extracellular siderophore production is required for full virulence [14] . In contrast , deletion of the SID1 gene in U . maydis does not affect virulence in maize [26]; instead , reductive iron assimilation is required for virulence [27] . This also true in Candida albicans , where ferric reductase activity is required for systemic infection [25] , [28] . Ferric reductase activity has been clearly demonstrated in H . capsulatum [8] , though the relevant genes that encode these activities have not been identified , and thus their role in virulence has not been assessed . Perusal of the genome revealed seven putative ferric reductase genes , though none of these candidates were observed to be upregulated by iron limitation in our experiments . Nonetheless , any or all of these genes could contribute to pathogenesis in the host . In addition to identifying a role for siderophore production in virulence , this study revealed several interesting regulatory properties of siderophore-biosynthesis genes . First , inspection of the sequences upstream of each gene revealed a consensus sequence that was present at least once per gene . The expression of several fungal orthologs of SID1 is repressed by GATA-type negative regulators that recognize the HGATR motif [19] , [20] , [29] . Interestingly , the consensus sequence we identified , 5′- ( G/A ) ATC ( T/A ) GATAA-3′ , contained an HGATAR motif , and was shown to bind an H . capsulatum GATA factor in vitro ( Chao et al . , submitted ) . We are now investigating whether these regulatory sequences are necessary to confer gene regulation in response to iron levels in vivo . Second , the siderophore biosynthesis genes were located adjacent to each other in the genome , and thus comprise the first secondary metabolite gene cluster defined in H . capsulatum . We observed similar clustering in the genome of the closely related systemic dimorphic fungal pathogen C . immitis , but it is not evident to nearly the same extent for siderophore biosynthesis genes in other sequenced fungal genomes [27] , [30] , [31] . Third , we could not express significant levels of SID1 from an episomal plasmid , suggesting that integration of the gene may be critical for normal expression levels . Additionally , integration of SID1 at random sites in complementation strains resulted in partial complementation of siderophore production , suggesting that optimal function might be achieved only with expression from the original genomic locus . Unfortunately , targeted integration is extremely challenging in H . capsulatum , making it difficult to directly test this hypothesis . The potential significance of the H . capsulatum gene cluster with regards to gene regulation is intriguing . One possibility is that clustering in the genome facilitates local changes in chromatin structure that allow a regional change in promoter accessibility . This type of regulation is reminiscent of transcriptional control of secondary metabolite clusters in Aspergillus species , where transcriptional accessibility of genes in the cluster is controlled by the putative methyltransferase LaeA [32] , [33] . Perhaps local control of chromatin structure allows a rapid and coordinated transcriptional switch in response to changes in iron levels .
Histoplasma capsulatum strains G217B ( ATCC 26032 ) and G186ARura5Δ ( WU8 ) , all kind gifts from the laboratory of William Goldman , Washington University , St . Louis , as well as strains generated in this study ( Table 1 ) were grown in HMM broth or plates [34] , or in mRPMI broth [RPMI 1640 medium without phenol red or bicarbonate ( Invitrogen , www . invitrogen . com ) , supplemented with 1 . 8% dextrose ( Fisher Scientific , www . fishersci . com ) , 25 mM HEPES pH 6 . 5 or pH 7 . 5 ( Fisher Scientific ) , and 100 units/mL of both penicillin and streptomycin ( UCSF Cell Culture Facility , www . ccf . ucsf . edu ) , modified from [8]] . Unlike HMM , mRPMI contained no added iron , but did not eliminate trace amounts of iron , though cultures grown in mRPMI medium were incubated in plastic flasks to reduce trace iron contamination . Media was supplemented with various concentrations of FeSO4 ( Fisher Scientific ) as described in the text . HMM and mRPMI media were supplemented when needed with 200 µg/mL hygromycin ( Roche , www . roche . com ) and 200 µg/mL uracil ( Sigma-Aldrich , www . sigmaaldrich . com ) . Cultures were grown at 37°C under 5% CO2 . For microarray studies , an initial culture of G217B yeast cells was grown in 5 mL HMM , and then passaged 1∶25 into 80 mL HMM . After 3 days of growth , the culture was pelleted , washed in 80 mL of PBS , and resuspended in 1 L of mRPMI pH 7 . 5 supplemented with 5 µM FeSO4 . After 24 hours of growth , 200 mL of culture was harvested for the zero time point . Then the culture was split into 2×400 mL , and an additional 5 µM or 10 µM FeSO4 was added to one culture and 100 µM deferoxamine mesylate ( Sigma-Aldrich ) was added to the other . At each time point , 100 mL of culture was harvested and processed as below . For growth , siderophore production , and QRT-PCR assays in G186AR based strains , cells were grown in 5 mL HMM . When the cultures reached late log phase , they were sonicated twice for 3 seconds to disperse cells and then passaged 1∶25 into HMM . After 24 hours of growth , they were pelleted , washed in an equal volume of PBS , and resuspended in an equal volume of mRPMI pH 6 . 5 . For growth curves , triplicate samples of 1 mL of cells were taken at each time point , sonicated twice for 3 seconds , and dilutions were measured by spectrophotometer at OD600 . For QRT-PCR , after resuspension in mRPMI pH 6 . 5 , cells were grown for an additional 24 hours . At that point , a t = 0 sample was harvested , then the culture was split . One culture was maintained in mRPMI pH 6 . 5 media with no additions , the other was treated with 5 µM FeSO4 . Cells were harvested at 1 , 4 , and 24 hours and total RNA was isolated using a guanidine thiocyanate lysis protocol as previously described [10] . Approximately 100 ng of PacI-linearized plasmid DNA with exposed telomere ends was transformed into yeast cells as previously described [35] . H . capsulatum yeast cells were transformed using Agrobacterium-mediated gene transfer as described previously [36] . Briefly , the A . tumefaciens strain ( LBA1100 , a kind gift of Thomas Sullivan and Bruce Klein with permission from Paul Hooykas ( Leiden University , Leiden , The Netherlands ) ) transformed with the desired plasmid was induced overnight with 200 µM acetosyringone ( AS , from Sigma-Aldrich ) . H . capsulatum yeast cells were harvested from 4 day patches on HMM+uracil plates and diluted to 5×108 cells/mL . Equal volumes of the H . capsulatum and A . tumefaciens cultures were mixed , and 400 µL of the mix was spread onto BiodyneA nylon membranes ( Pall Gelman , www . pall . com ) on IM agarose plates containing 200 µM AS and 200 µg/mL uracil . Co-cultivation plates were incubated at 28°C for 3 days . The membranes were then transferred onto HMM plates with no added uracil and incubated at 37°C for 2 to 3 weeks . Strains WU8 , HcLH25 , and HcLH26 were transformed with pVN61 ( vector control marked with URA5 ) and pED2 ( SID1 complementation marked with URA5 ) . All of the transformed strains show similar in vitro growth in liquid and solid media without uracil . Cultures of H . capsulatum were harvested by filtration , and total RNA was isolated using a guanidine thiocyanate lysis protocol as previously described [10] . This RNA was used for both microarrays as well as QRT-PCR . For microarrays , polyadenylated RNA was purified from total RNA using an Oligotex mRNA kit ( Qiagen Inc . , www . qiagen . com ) . cDNA synthesis from polyA-selected RNA and fluorescent labeling was performed as described previously [37] . Briefly , an equal mass of RNA from each time point was pooled to generate a reference sample and labeled with Cy3 . cDNA from each individual time point was labeled with Cy5 and competitively hybridized versus the reference pool using H . capsulatum G217B 70-mer oligonucleotide microarrays . The microarrays contain a single 70-mer oligo for each predicted gene in the G217B genome , as well as two 70-mer oligos for low confidence genes . Arrays were scanned on a GenePix 4000B scanner ( Axon Instruments/Molecular Devices , www . moleculardevices . com ) and analyzed using GENEPIX PRO , version 6 . 0 ( Molecular Devices ) , Spotreader ( Niles Scientific , www . nilesscientific . com ) , NOMAD 2 . 0 ( http://nomad2 . ucsf . edu/NOMAD/nomad-cgi/login . pl ) , CLUSTER [38] , and Java Treeview 1 . 0 . 12 ( available at http://sourceforge . net/project/showfiles . phpgroup_id84593 ) . To eliminate elements with low signal , we did not analyze elements for which the sum of the medians for the 635-nm and 532-nm channels was ≤500 intensity units . All of the time points were normalized relative to the zero time point . Total RNA was treated with DNaseI ( Promega , www . promega . com ) . cDNA was synthesized from 3 . 3 µg of DNaseI-treated RNA using Stratascript reverse transcriptase ( Stratagene , www . stratagene . com ) and oligo-dT . Quantitative PCR was performed on 1∶100 dilutions of cDNA , except for MFS2 and SID5 reactions , which used 1∶40 dilutions of cDNA . The reactions included 1 . 5 mM MgCl2 , 1× Amplitaq buffer , 0 . 6 units Amplitaq Gold ( Applied Biosystems , www . appliedbiosystems . com ) , 1× SYBR Green ( Molecular Probes , probes . invitrogen . com ) , and 200 nM primers . Reactions were performed on the Mx3000P QPCR system ( Stratagene , www . stratagene . com ) with Comparative Quantitation ( with dissociation curve ) program , using actin ( ACT1 ) as the normalizing transcript . Cycling parameters were 95°C for 10 minutes , then 40 cycles of 95°C ( 30 s ) , 57°C ( 1 min ) , 72°C ( 30 s ) followed by dissociation curve analysis . All reactions were performed in triplicate . A calibrator sample made from an equal mix of each RNA sample was included on each plate . Reactions were analyzed using MxPro software . Primer sequences are included in supplemental material ( Table S1 ) . The H . capsulatum GenBank/NCBI ( http://www . ncbi . nlm . nih . gov/ ) nucleotide sequences for the genetic loci described in this publication are ABC1 ( EU253969 ) , MFS1 ( EU253970 ) , MFS2 ( EU253971 ) , NIT22 ( EU253972 ) , NPS1 ( EU253973 ) , OXR1 ( EU253974 ) , RTA1 ( EU253975 ) , SID1 ( EU253976 ) , SID3 ( EU253977 ) , SID4 ( EU253978 ) , and SID5 ( EU253979 ) . Promoter regions ( including 1 kb upstream of ATG , or entire intergenic regions for SID4/SID1 and NPS1/ABC1 ) of SID3 , SID4 , SID1 , NIT22 , NPS1 , ABC1 , OXR1 , and MFS1 from G217B and G186AR were submitted for MEME ( Multiple Em for Motif Elicitation ) analysis at http://meme . sdsc . edu . The length range for consensus site was set between 6 and 50 base pairs . A positive-negative selection strategy was utilized to disrupt SID1 in G186ARura5Δ ( WU15 ) , as described in [39] . Originally , we attempted to disrupt this gene in the G217Bura5-23 strain , but we were unsuccessful . A disruption construct , pLH36 , was made containing 1213 bps 5′ and 943 bps 3′ of the SID1 open reading from G217B . The open reading frame was replaced with the hygromycin resistance gene , hph under the control of the Aspergillus nidulans gpd promoter . This construct was introduced into G186ARura5Δ ( WU15 ) by electrotransformation . Seven independent transformants were colony purified , inoculated into HMM+hygromycin+uracil+20 µM FeSO4 medium and passaged 3 times ( 1∶25 dilution ) once the cultures had reached mid log phase . After the final passage , serial dilutions of the cells were plated onto HMM ( pH 4 . 5 ) +1 g/L 5-Fluoroorotic acid ( 5-FOA , Zymo Research , www . zymoresearch . com ) +20 µM FeSO4 agarose plates . Genomic DNA from colonies on 5-FOA plates was tested by PCR for the disruption . Potential gene disruptions were further tested by Southern analysis . Secretion of siderophores was detected using chrome azurol S ( CAS ) [40] . Briefly , 0 . 5 mL of culture supernatant was mixed with 0 . 5 mL of CAS assay solution ( 600 µM hexadecyltrimethyl ammonium ( Sigma Aldrich ) , 15 µM FeCl3 ( Sigma Aldrich ) , 150 µM CAS ( Sigma Aldrich ) , 500 mM anhydrous piperazine ( Fluka ) , 750 mM HCl ( Fisher Scientific ) , and 4 mM 5-sulfosalicylic acid ( Sigma Aldrich ) ) , incubated for 1 hour , and the OD630 was measured with a spectrophotometer . Bone-marrow-derived macrophages ( BMDM ) were isolated from femurs of 6- to 8-week old female C57BL/6J mice . The bone marrow was eluted using BMM ( Bone Marrow Macrophage ) medium , which consists of Dulbecco's Modified Eagle Medium , D-MEM High Glucose ( UCSF Cell Culture Facility ) , 20% Fetal Bovine Serum ( Hyclone , Thermo Fisher , www . hyclone . com ) , 10% v/v CMG supernatant ( the source of CSF-1 ) , 2 mM glutamine ( UCSF Cell Culture Facility ) , 110 µg/mL sodium pyruvate ( UCSF Cell Culture Facility ) , penicillin and streptomycin ( UCSF Cell Culture Facility ) . The bone marrow cells were cultured for 6 days at 37°C with 5% CO2 . The cells were harvested with cold PBS ( without Mg and Ca ) , and frozen in BMM+10% DMSO . For infections , H . capsulatum cells were grown to late log phase in HMM . The cells were pelleted and resuspended in BMM medium , sonicated twice for 3 seconds , and counted by hemacytometer . Approximately 1×106 H . capsulatum cells were used to infect 2×105 BMDM in 24-well cell culture dishes . After a 1 hr incubation , the infected macrophages were washed twice with D-MEM High Glucose and then incubated in 500 µL fresh BMM medium with or without 100 µM FeSO4 at the zero time point . At 6 , 24 and 48 hrs , the medium was removed from the wells and 500 µl H2O was added . After 5 min incubation , the lysed macrophages were transferred to a 1 . 5 mL eppendorf and sonicated for 3 sec . Dilutions were plated onto HMM plates to determine CFUs . CFUs were normalized to the zero time point . Significance was determined using ANOVA with Bonferroni Multiple Comparisons Test . H . capsulatum strains hcLH95 ( wild-type ) , hcLH97 ( sid1Δ::hph ) , and hcLH106 ( sid1Δ::hph+SID1 ) were grown in 5 mL of HMM . Cells were passaged once and grown until late log phase . 1 mL of culture was sonicated twice for 3 seconds to fully disperse cells then washed twice with cold PBS . Cell concentration was determined by hemacytometer . Mice were infected intranasally with 5×104 cells in 25 µL PBS containing an equal proportion of hcLH95 and hcLH97 ( wild-type and sid1Δ::hph ) or hcLH95 and hcLH106 ( wild-type and sid1Δ::hph+SID1 ) followed by 5 µL of PBS chase . Four mice were used for each time point . Lungs were homogenized in 5 mL of HMM . Serial dilutions were plated onto both HMM and HMM+hygromycin plates in order to distinguish sid1Δ::hph and sid1Δ::hph+SID1 strains from the wild-type strain . Enumeration of wild-type , sid1Δ , and sid1Δ+SID1 yeast allowed for the determination of competitive index ratio ( CI ) using the following formula: CI = ( mutant output/wild-type output ) / ( mutant input/wild-type input ) . Input was determined at four hours post infection . Significance of the results was determined with the Kruskal-Wallis Test ( ANOVA ) with Dunn's Multiple Comparisons Test and with the Mann-Whitney/Wilcoxon rank sum test . Enumeration of yeast by patch test onto HMM and HMM+hygromycin was also performed and yielded a CI identical to the plating of serial dilutions ( data not shown ) . Female C57BL/6J mice were purchased from The Jackson Laboratories . Experiments were in accordance with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at U . C . San Francisco . | Fungal infections are a growing public health threat , particularly for immunocompromised individuals such as people with AIDS , organ transplant recipients , and cancer patients . Present antifungal therapies are often highly toxic and resistance to these therapies continues to rise . Histoplasma capsulatum is a pathogenic fungus that infects humans , causing pulmonary and systemic disease . It is the most common cause of fungal respiratory infection in the world , and is endemic to the Mississippi and Ohio River valleys of the United States . H . capsulatum produces small molecules , called siderophores , to acquire iron , an essential nutrient . We have identified genes that are involved in the synthesis of siderophores in this fungus and have found that siderophore production in H . capsulatum is important for its virulence . Since siderophore production is confined to microbes and plays no role in human biology , it is an excellent target for rational drug design . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/fungal",
"infections",
"microbiology/microbial",
"physiology",
"and",
"metabolism",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis"
] | 2008 | Histoplasma Requires SID1, a Member of an Iron-Regulated Siderophore Gene Cluster, for Host Colonization |
Host genetic variations play an important role in several pathogenic diseases , and we have previously provided strong evidences that these genetic variations contribute significantly to differences in susceptibility and clinical outcomes of invasive Group A Streptococcus ( GAS ) infections , including sepsis and necrotizing soft tissue infections ( NSTIs ) . Our initial studies with conventional mouse strains revealed that host genetic variations and sex differences play an important role in orchestrating the severity , susceptibility and outcomes of NSTIs . To understand the complex genetic architecture of NSTIs , we utilized an unbiased , forward systems genetics approach in an advanced recombinant inbred ( ARI ) panel of mouse strains ( BXD ) . Through this approach , we uncovered interactions between host genetics , and other non-genetic cofactors including sex , age and body weight in determining susceptibility to NSTIs . We mapped three NSTIs-associated phenotypic traits ( i . e . , survival , percent weight change , and lesion size ) to underlying host genetic variations by using the WebQTL tool , and identified four NSTIs-associated quantitative genetic loci ( QTL ) for survival on mouse chromosome ( Chr ) 2 , for weight change on Chr 7 , and for lesion size on Chr 6 and 18 respectively . These QTL harbor several polymorphic genes . Identification of multiple QTL highlighted the complexity of the host-pathogen interactions involved in NSTI pathogenesis . We then analyzed and rank-ordered host candidate genes in these QTL by using the QTLminer tool and then developed a list of 375 candidate genes on the basis of annotation data and biological relevance to NSTIs . Further differential expression analyses revealed 125 genes to be significantly differentially regulated in susceptible strains compared to their uninfected controls . Several of these genes are involved in innate immunity , inflammatory response , cell growth , development and proliferation , and apoptosis . Additional network analyses using ingenuity pathway analysis ( IPA ) of these 125 genes revealed interleukin-1 beta network as key network involved in modulating the differential susceptibility to GAS NSTIs .
The human skin acts as a barrier between the external environment and the body , protecting it from pathogens and regulating body temperature [1] . Nevertheless , skin infections occur . Group A Streptococcus ( GAS ) or Streptococcus pyogenes is one of the most common causative agents of human skin and severe soft tissue infections , which can range from uncomplicated impetigo and cellulitis to life-threatening necrotizing soft tissue infections ( NSTIs ) [2–9] . Our previous clinical investigations have shown that globally disseminated M1T1 GAS isolates can be isolated from patients with pharyngitis as well as from patients with severe invasive NSTIs , emphasizing the host factor contributions to the outcomes of invasive GAS infections [10] . Indeed , later studies revealed that allelic variations in human leukocyte antigen ( HLA ) class II haplotypes resulted in striking differences in severity and outcomes of invasive GAS infections via the differential presentations of GAS superantigens ( SAgs ) by class II HLAs to host T-cell receptor ( TCR ) Vβ elements; this finding further revealed a strong role for host factors in modulating GAS disease outcomes [11–14] . Nevertheless , in addition to SAgs , GAS harbors a wide array of virulence factors , including antiphagocytic M protein , fibronectin-binding proteins , proteases , DNases , and haemolysins , that can interact with a wide array of host factors besides HLA class II molecules during GAS pathogenesis . These unknown factors can influence susceptibility and severity of GAS NSTIs [15] . Although GAS causes NSTIs in humans , successful animal models have been developed through the use of conventional inbred and outbred mouse strains; however , mice are generally more resistant to GAS NSTIs than humans and thus require higher initial GAS inoculum than humans to develop NSTIs [16–23] . Our initial studies to understand the role of host genetic context in GAS NSTIs revealed that the DBA/2J ( D2 ) mouse strain is more susceptible than is the C57BL/6J ( B6 ) strain and that sex differences have a possible role in potentiating the severity of NSTIs [24] . However , the restricted genetic variations in these inbred strains led us to develop a better animal model for use in further in-depth investigations of the host genetic architecture for GAS NSTIs susceptibility . For this purpose and to identify additional host genetic and nongenetic factors , we undertook our present study utilizing an advanced recombinant inbred ( ARI ) mouse strains ( BXD ) . We applied a forward systems genetics approach to map quantitative trait loci ( QTL ) harboring genes that are most likely to be involved in modulating susceptibility/outcomes in a NSTIs model . A panel of distinct yet replenishable , genomically defined , and fully genotyped ARI mouse strains ( BXD ) [25] , which were derived from B6 and D2 strains , were infected with an equal dose of M1T1 GAS bacteria [2 , 10] and analyzed to detect variations in disease severity phenotypes . The quantitative phenotypes measured included survival , weight change kinetics , lesion size , bacterial load , and dissemination to organs . In summary , we have established and optimized a murine model of NSTIs by utilizing a well-characterized and widely disseminated strain of GAS . Using this model , we mapped four NSTIs-associated QTL regions in the mouse and identified 125 polymorphic differentially expressed genes in representative susceptible BXD strains that have a high likelihood of modulating susceptibility to NSTIs . Finally , using network analyses , we narrow down interleukin-1 beta ( IL-1β ) as a key regulator in modulating NSTIs susceptibility .
Our laboratory’s preliminary studies of subcutaneous GAS infections of conventional mouse strains determined that D2 mice are more susceptible to GAS NSTIs than are B6 mice; also D2 female mice are more resistant than are males of the same strain [24] . To further explore the relative contributions of host genetics , sex , and any other covariates to GAS NSTIs , we infected 33 BXD strains along with their ancestral parental strains ( B6 and D2 ) and their F1 strains ( B6D2F1 ) with an equal dose of GAS bacteria . Confirming our prior observations , the susceptibility of D2 mice was greater than that of B6 mice ( Fig 1 ) ; also D2 female mice and female mice of other BXD strains were resistant in terms of survival ( Fig 2 ) . To further delineate the host genetic architecture of GAS NSTIs , we monitored three quantifiable phenotypic traits in mice of these infected strains: survival , weight change , and lesion size . To identify significant QTL that modulate the observed phenotypes , we performed genome-wide linkage scans ( GWLS ) by mapping quantitative phenotypic traits to the BXD mouse genotypes available from the Gene Network ( GN ) . To validate our mapped survival QTL , we selectively infected a BXD strain with a D ( or B ) haplotype within the observed QTL region and determined whether such mice exhibited a susceptible phenotype such as increased mortality associated with increased bacterial burden or a resistant phenotype such as reduced mortality associated with reduced bacterial burden . Based on the mapping of the survival phenotype to the QTL ( Fig 5B ) , we chose 10 BXD strains with either a B or a D haplotype between 24 . 5 and 35 Mb at Chr 2 and infected them and their ancestral parents ( B6 and D2 ) with an equal dose of GAS bacteria . All the BXD strains with B haplotypes within the survival QTL ( and the B6 strain ) were resistant in terms of survival ( Fig 9A ) , whereas the BXD strains with D haplotypes ( and the D2 strain ) were found to be differentially susceptible in terms of survival ( Fig 9B ) . These differences in survival confirmed the role of other confounding nongenetic cofactors ( including sex , age , and body weight ) . Next , we looked for any marked differences in the bacterial load in skin and blood and in bacterial dissemination between the BXD mice with either the B or the D haplotypes within the survival QTL . All BXD strains with D haplotypes and the D2 strain contained significantly greater bacterial loads in skin , greater bacteremia , and greater dissemination to the spleen than did the BXD strains with B haplotypes and the B6 strain ( Fig 10A–10C ) . PCA of these three non-independent bacterial load measurements revealed one major component , PC1 , which covered 95 . 2% of the variances ( Fig 10D ) . The BXD strains were grouped together on the basis of their haplotypes , with the BXD strains with a D haplotype representing a positive group in terms of increased bacterial load and dissemination . GWLS of three monitored phenotypic traits relevant to NSTIs ( survival , percent weight change , and lesion size ) revealed four QTL that contain 516 genes on mouse Chr 2 ( 276 genes ) , 7 ( 98 genes ) , 6 ( 106 genes ) , and 18 ( 36 genes ) . From this extensive gene list , we wanted to further narrow down and identify potential host candidate genes . First , we used the QTLminer tool available in GN [28] to investigate our QTL and rank-order the genes ( rank 0–4 ) based on ( i ) gene annotation data ( including gene name , description , Gene Ontology terms ) , ( ii ) normal gene expression datasets ( adipose , muscle , and spleen mRNA ) , ( iii ) gene polymorphism data ( non-synonymous single nucleotide polymorphisms [nsSNPs] found between the ancestral parental B6 and D2 strains and insertion or deletion of bases [INDELs] found in the BXD strains ) , and ( iv ) genes that have local genetic control ( cis-eQTLs ) . From these rank-ordered genes , we shortlisted 375 host candidate genes on mouse Chr 2 ( 224 genes ) , 7 ( 74 genes ) , 6 ( 55 genes ) , and 18 ( 22 genes ) on the basis of their annotation data and biological relevance to GAS NSTIs ( S1–S4 Tables ) . Next , these 375 shortlisted genes were further evaluated for their differential expression profiles at 48h post-infection , between the representative susceptible ( BXD 40 and 64 ) and resistant ( BXD 73 and 87 ) BXD strains compared with their respective uninfected controls , using custom designed 384-well real time PCR plates with PrimePCR SYBR Green assays as described in the methods section . Accordingly , we found that 125 genes were significantly ( FDR < 0 . 10 ) differentially expressed ( ±1 . 5 fold regulation as threshold ) in susceptible strains . However , none of the tested genes were significantly differentially regulated in resistant strains compared to their uninfected controls . Detailed lists of all these genes along with their expression fold regulation values are given in S5–S7 Tables . Finally , we parsed these significantly ( FDR < 0 . 10 ) differentially expressed ( ±1 . 5 fold regulation as threshold ) 125 genes into pathways using Ingenuity Pathway Analysis ( IPA ) . Through this , we extracted two high-scoring networks , which are shown merged in Fig 11 , comprising genes that are related to cancer , cell death and survival , inflammatory response , organismal injury and abnormalities , and tissue morphology . Based on these networks , interleukin-1 β ( IL-1β ) was likely to be the key response molecule regulating susceptibility to GAS NSTIs . We then tested its differential regulation at 48h post-infection , in both ancestral strains ( B6 and D2 ) and between the representative BXD strains ( BXD 69 , 99 , 100 and 102 ) compared with their respective uninfected controls . These BXD strains were selected based on their respective cRSI ( survival index ) as shown in Fig 1 and covered a wide range of survival indices between the two ancestral strains . As expected , we observed that IL-1β was significantly up regulated in D2 parental strains compared to B6 ( Fig 12A ) . In addition , we observed that the IL-1β expression was significantly correlated with infection severity as shown in Fig 12B , wherein we found that strains with lower cRSI values ( decreased survival ) had significantly higher IL-1β expression compared to their respective uninfected controls and vive versa . To further validate IL-1β as a potential key factor involved in susceptibility to GAS NSTI , we conducted a proof-of-concept experiment in which plasma samples and tissue biopsies from an M1 GAS infected NSTI patient ( 2006 ) , as well as samples from an in vitro experimental model of human skin infected with the patient’s isolate were analyzed . Light microscopy analysis of Gram-stained tissue biopsies collected on days 1 and 2 identified Gram-positive cocci in the affected tissue with a massive bacterial load noted in the day 1 biopsy while substantially fewer cocci were detected in the day 2 biopsies ( Fig 13A and 13B ) . Analyses of IL-1β gene expression showed an increased expression in patient biopsies as compared to biopsies from healthy controls ( Fig 13C ) . In line with the bacterial load , IL-1β gene expression decreased in the day 2 biopsy . Similarly , analyses of systemic IL-1β revealed high levels in patient plasma collected day 0 while the day 3 sample had negligible amounts close to that of healthy controls ( Fig 13D ) . In order to study the IL-1β response in an experimental model , we utilized a three dimensional ( 3D ) tissue model of human skin [29] , which was infected with GAS strain 2006 . Gram staining of infected tissue models revealed bacterial dissemination throughout the entire tissue already 24h after infection and similarly to the patient findings , high bacterial load was evident at 48h after infection ( Fig 13E ) . Furthermore , analysis of tissue model supernatants revealed increasing IL-1β levels over the infection period , whereas in unstimulated models only background levels were detected ( Fig 13F ) .
The ARI panel of BXD mouse strains has been widely used to identify the host genetic loci for pathogenic disease susceptibility [26 , 27 , 30–33] . The major advantage of using this mouse model in the study of complex host-pathogen interactions is that the genomes of the two ancestral parents of the BXD strains , namely B6 and D2 , differ from each other by approximately 1 . 8 million SNPs , and thus the genome of each fully genotyped , renewable BXD strain has a unique recombination mixture of chromosomal segments inherited from either parent [25 , 34] . In this study , we utilized 33 such BXD strains ( along with the B6 , D2 , and B6D2F1 strains ) and mapped three quantitative phenotypic traits—survival , weight change , and lesion size , to define the genetic architecture of GAS NSTIs in the mouse . In addition to strain variability ( host genetic context ) , nongenetic cofactors including sex , age , and body weight are also significant predictors in controlling the susceptibility and severity of GAS NSTIs . The major findings of this study include the identification of four QTL on mouse chromosomes ( Chr 2 , 6 , 7 , and 18 ) and the identification of 375 host candidate genes ( S1–S4 Tables ) that have a likely role in dictating severity , susceptibility , and manifestations of GAS NSTIs . Further differential expression analyses associated interleukin-1 beta pathway as key network involved in regulating GAS NSTIs severity and susceptibility . In our initial studies with conventional inbred and outbred mouse models [24] and in this study , the parental strain D2 was more susceptible to GAS NSTIs than was the B6 strain , but interestingly the resistant B6 strain lost more weight and developed larger lesions than did D2 . Susceptibility , in terms of the three phenotypic traits based on their relative genotypes , differed greatly among the BXD strains . For instance , BXD strains with D haplotypes within the survival QTL on mouse Chr 2 were susceptible to NSTIs compared to BXD strains harboring B haplotypes . Further , BXD strains with D haplotypes within the lesion QTL on Chr 6 and B haplotypes within the weight change QTL on mouse Chr 7 and/or lesion size QTL on Chr 18 lost the most weight and developed large lesions respectively . Based on these findings , we can predict the relative severity and/or manifestations of GAS NSTIs in any of the BXD strains . For instance , BXD 40 was an extremely susceptible strain in our study; retrospectively , we associated susceptibility and severity with the presence of B haplotypes within the QTL on Chr 7 ( increased weight loss ) and on Chr 18 ( large lesions ) and the D haplotypes on the QTL on Chr 6 ( large lesions ) and Chr 2 ( increased mortality , bacterial load , and dissemination ) . We propose the collective polygenic contributions of these four QTL enabled BXD 40 to be more susceptible than the D2 parent . In contrast , like the resistant B6 parents , BXD 73b and 87 strains harbor B haplotypes on the Chr 18 QTL ( lesion size ) and on the Chr 2 QTL ( increased survival with reduced bacterial load and dissemination ) ; therefore , these strains are resistant to GAS NSTIs and recover if such infection occurs . However , the additional D haplotypes on Chr 6 QTL ( large lesions ) may be associated with both the BXD strains developing larger lesions , whereas the incomplete association of B haplotypes on proximal Chr 7 QTL ( increased weight loss ) may be the reason why BXD 87 lost less weight than the B6 parental strain did . Our prior studies using the BXD strains to define the genetic susceptibility loci for survival against GAS sepsis mapped to three QTL , including a strong QTL between 24 and 34 Mb on mouse Chr 2 that had a peak LRS of 34 . 2 ( GN trait ID: 10836 ) [30] . Although both studies were independent and involved different forms of invasive M1T1 GAS isolates ( 5448WT for NSTIs vs . 5448AP for sepsis ) , which were administered by two different routes ( subcutaneous vs . intravenous tail vein injections ) , we located one overlapping QTL on mouse Chr 2 for survival in both studies . In fact , 5448AP is an isogenic , animal-passaged variant of the clinical isolate 5448WT , which exhibits a hypervirulent phenotype and has an expression profile of virulence factors that differs from that of its parental strain . The main advantage of 5448AP is that it lacks the cysteine protease SpeB , which can degrade host and pathogenic proteins without discrimination , and in the absence of SpeB , 5448AP preserves most of its virulent signature proteins and thereby evades the host immune strategies [35–39] . Host signals including those of neutrophils [40] , transferrin and lactoferrin [41] , and transition metals ( zinc and copper ) [42] have been shown to facilitate the development of a SpeB-negative phenotype in 5448WT . However , SpeB is needed for skin infections [43] , and that was one of the reasons we used 5448WT for our GAS NSTIs studies . Taken together , we speculate that the 5448WT strain produces SpeB needed for initial NSTI pathogenesis , but once the pathogen reaches deep layers of skin tissue and enters the bloodstream , it may become a hypervirulent variant ( 5448AP ) that causes sepsis in addition to NSTIs . For this reason , we believe the two independent BXD studies ( sepsis and NSTIs ) studying the genetic susceptibility loci for survival showed an overlap in QTL on mouse Chr 2 . We are currently exploring this scenario in depth to understand the exact spatio-temporal transition from WT to AP phenotype and the molecular host-pathogen interactions that lead to this transition . In addition , we are determining whether host genetics has a role in influencing this transition by mapping the transition ratio ( ratio of SpeB-negative phenotypes in the total bacteria isolated from the infected skin and/or organs of each BXD mouse at specific time points after infection ) as a quantitative phenotype . In this way , we will define the QTL encompassing the candidate genes influencing this transition . We used the QTLminer tool to explore the four mapped host susceptibility loci and extracted 375 rank-ordered , biologically relevant host candidate genes , most of which are polymorphic in both ancestral strains ( B6 and D2 ) and are involved in inflammation , innate immunity , cell cycle , and apoptosis among other biological processes ( S1–S4 Tables ) . Our differential expression analyses of these candidate genes between the representative resistant ( BXD 73 and 87 ) and susceptible ( BXD 40 and 64 ) BXD strains before and after GAS NSTIs further narrowed the list of candidate genes . Specifically , 125 genes were significantly differentially expressed in susceptible strains compared to their uninfected controls ( S5–S7 Tables ) . Significantly down regulated genes on mouse Chr 2 ( trait: survival ) and Chr 6 ( trait: lesion size ) included abelson murine leukemia oncogene 1 ( Abl1 ) , adenylate kinase 1 ( Ak1 ) , anaphase promoting complex subunit 2 ( Anapc2 ) , endothelial differentiation-related factor 1 ( Edf1 ) , ets variant gene 6 ( Etv6 ) , LIM homeobox transcription factor 1 beta ( Lmx1b ) , netrin G2 ( Ntng2 ) , nucleoporin 214 ( Nup214 ) , notch gene homolog 1 ( Notch1 ) , retinoid X receptor alpha ( Rxra ) and tuberous sclerosis 1 ( Tsc1 ) ; each is involved in cell cycle arrest , cell proliferation and differentiation , angiogenesis , and apoptosis as indicated by the gene ontology ( GO ) annotation data provided by QTLminer ( S1–S7 Tables ) . Further studies are needed to elucidate if and how these observed quantitative differences relate to differential susceptibility to GAS NSTIs . On the other hand , expression of the gene for glycerophosphodiester phosphodiesterase 1 ( Gde1 ) on Chr 7 ( trait: weight change ) has been associated with triglyceride accumulation in mice [44] and in our case Gde1 was significantly down regulated in susceptible strains compared to their uninfected controls . We hypothesize that this may be one of the reasons for the observed weight loss in these two susceptible BXD strains ( Figs 3 and 7 ) . Our gene network analyses of these 125 genes further identified two high-scoring gene networks that likely play a significant role in determining susceptibility to GAS NSTIs , particularly highlighting interleukin-1 β ( IL-1β ) as key regulator likely to participate in modulating GAS NSTIs ( Fig 11 ) . IL-1β produced chiefly by monocytes , macrophages , and dendritic cells , belongs to the interleukin-1 family of cytokines , and is the highly inflammatory cytokine associated with several inflammatory conditions [45] . Indeed , our subsequent gene expression analyses revealed us that IL-1β was in fact significantly up regulated in the susceptible D2 parental strains compared to B6; also , there was a significant correlation between infection severity ( as measured by survival index ) and IL-1β expression levels ( Fig 12 ) . We also observed IL-1β to be significantly up regulated in patient biopsies as compared to biopsies from healthy controls ( Fig 13C ) . Besides , the protein levels of IL-1β were also increased in patient plasma samples and in infected skin models compared to their respective controls ( Fig 13D and 13F ) , suggesting that GAS NSTIs susceptibility is associated with an amplified inflammatory response mediated by IL-1β . However , although IL-1β-dependent macrophage immune response has been shown to be protective in an intravenous mouse model of GAS sepsis utilizing 5448AP bacteria [46] , female CD-1 strains were used for these infections and we have shown that these strains are resistant to GAS NSTIs mediated by 5448WT bacteria [24] . Moreover , similar to what we mentioned earlier , both these studies use two different infection models ( sepsis vs . NSTIs ) and bacteria ( 5448AP vs . 5448WT ) . Taken together , we believe that host genetics- and context-dependent differential regulation of IL-1β can modulate the severity and mortality of GAS invasive infections . Further investigations are warranted to have a better understanding of this phenomenon . In conclusion , our unbiased forward systems genetics approaches utilizing the ARI panel of BXD mice strains has defined the genetic architecture of GAS NSTIs in mice revealing four NSTIs-associated QTL on mouse Chr 2 , 6 , 7 and 18 , and has allowed us to extract 375 host candidate genes . Additional differential expression analyses between the representative resistant ( BXD 73 and 87 ) and susceptible ( BXD 40 and 64 ) BXD strains , before and after GAS NSTIs , allowed us to identify 125 significantly differentially expressed genes in susceptible strains compared to their uninfected controls . Our IPA analyses of these 125 genes underscored IL-1β as potentially a key regulator of GAS NSTIs susceptibility; a finding that was also supported by its up regulation in patient tissue biopsies as well as in GAS infected experimental models including mice and engineered 3D human skin tissue . Our ongoing in-depth pathway analyses to identify additional molecular interactions between all the differentially expressed genes will help us dissect various host mechanisms and/or interactions with GAS resulting in the differential susceptibility to NSTIs .
For the NSTI studies , we used the ARI lines ( BXD ) [25] , their ancestral parental strains ( C57BL/6J [B6] and DBA/2J [D2] ) , and their F1 population ( B6D2F1 ) . All mouse strains were obtained from the in-house breeding colonies at UC and UND . A total of 629 mice were used; 508 mice ( 262 males and 246 females ) , after the exclusions based on previously described , predetermined criteria [30 , 31] were considered for final analyses . Tissue biopsies and plasma samples collected from one GAS NSTI patient , a 61 year old male ( patient ID 2006 ) , enrolled at Rigshospitalet in Copenhagen as part of the EU-funded project INFECT ( www . fp7infect . eu ) were analyzed . The tissue biopsies were collected from the site of infection , e . g . fascia ( day 1 ) and soft tissue ( day 2 ) , during surgical procedures and immediately snap-frozen . Skin tissue from healthy controls was obtained at plastic surgery at the Karolinska University Hospital . Also , plasma samples were obtained from healthy individuals . We used a representative M1T1 clonal GAS isolate 5448WT ( for animal studies ) and NSTI patient M1 GAS isolate 2006 ( for skin model studies ) , which were routinely grown at 37°C in THY medium ( Todd-Hewitt broth ( Difco ) supplemented with 1 . 5% ( w/v ) yeast extract ) statically as described previously [10] . The human keratinocyte cells ( N/TERT-1 ) were maintained in EpiLife medium ( Invitrogen ) . Normal human dermal fibroblasts were cultured in DMEM ( Invitrogen ) supplemented with 10% ( v/v ) fetal bovine serum ( FBS; Invitrogen ) . Both were cultured at 37°C under a 5% CO2 atmosphere . The models were generated following the protocol as previously published [29] and were infected with 1 x 106 CFU of M1 GAS isolate 2006 for 24 and 48h . Hair was removed from the dorsal side of the mouse with Nair hair removal lotion ( Church & Dwight Co . , Inc . , Ewing , NJ ) one day prior to infection . Groups of 5–36 mice from a total of 33 BXD strains , their ancestral parental strains ( B6 and D2 ) , and the B6D2F1 strains were infected subcutaneously under the skin with 1 x 108 CFU of 5448WT per mouse ( bacteria were suspended in 100 μL of sterile phosphate-buffered saline [PBS] per dose ) , while control animals were given injections of sterile PBS alone . After inoculation , each mouse was housed in individual cages to avoid contact/influence from other animals over lesions . Animals were observed twice a day for the next seven days for mortality , weight change , and lesions . For bacterial load estimation studies , animals were humanely sacrificed ( euthanized ) if needed during the seven-day infection timeline ( or at the experimental endpoint ) . Blood was drawn through cardiac puncture for bacteremia estimations; necrotic skin tissues and spleens were also recovered and homogenized by a rotor stator homogenizer ( Omni International , Marietta , GA ) for enumeration to determine bacterial load and dissemination , respectively . We calculated broad sense heritability as previously described with slight modifications [27 , 31] . We used the OLS ANOVA table computed by GLM analysis to calculate broad sense heritability , which was expressed as the ratio of genetic variance ( mean square of strain ) to total variance ( sum of mean square of strain and residuals ) . We performed genome-wide linkage scans ( GWLS ) to identify QTL by using a web-based QTL/interval mapping tool ( WebQTL ) available on the Gene Network ( GN ) website ( www . genenetwork . org ) . This tool scans the mouse genome for QTLs and estimates the strength of the linkage as likelihood ratio statistic ( LRS ) using 5000 permutation tests [34] . We performed three sets of GWLS by using strain means for the following three quantitative phenotypic variables , for which the original datasets can be obtained at GN by using the associated identification numbers: ( i ) corrected relative survival index ( GN Trait ID: 17524 ) , ( ii ) corrected percent weight change kinetics for days 1–4 and their principal component ( PC1 ) ( GN Trait ID: 17520–17523 and 17527 ) , and ( iii ) corrected maximum lesion area ( GN Trait ID: 17525 ) . We used the web-based QTLminer tool available in the GN website to rank-order and shortlist the host candidate genes on the mapped QTL [28] . Representative BXD strains selected based on their susceptibility to GAS NSTIs ( BXD 40 and 64 for susceptible strains , and BXD 73 and 87 for resistant strains ) was subcutaneously infected with either 1x108 GAS 5448WT or mock infected with PBS . Three independent experiments ( n ≥ 3 mouse per strain , per experiment ) were performed . Forty-eight hours after infection , animals were humanely sacrificed ( euthanized ) and necrotic skin tissues were excised from each of the mouse , from which total RNA was isolated using FastRNA PRO green kit ( MP Biomedicals , LLC , Santa Ana , CA ) , and then purified using GeneJET RNA cleanup and concentration micro kit ( ThermoFisher Scientific , Grand Island , NY ) . RNA samples ( with A260/280 ≥ 1 . 9 ) from same strain and sex were pooled together for cDNA synthesis by SensiFAST cDNA synthesis kit ( Bioline , Taunton , MA ) . We custom designed 384-well real-time PCR plates with 384 unique PrimePCR SYBR Green assays ( Bio-Rad , Hercules , CA ) with most gene specific primers that span long introns to distinguish cDNA from genomic DNA and utilized Bio-Rad CFX384 Touch real-time PCR detection system ( Bio-Rad , Hercules , CA ) . We used beta actin ( Actb ) , beta glucuronidase ( Gusb ) and ribosomal protein L7A ( Rpl7a ) as the endogenous control , which were used to normalize our gene expression data . Real-time PCR experiments often results in non-detects , where the PCR reactions failed to produce a minimum amount of signal due to various reasons . Studies have shown that imputing missing expression values contribute to less biased estimation of non-detects from real-time PCR experiments [47 , 48] . The PrimePCR analysis software ( Bio-Rad , Hercules , CA ) , used in this study to obtain relative normalized fold regulation and statistical significance ( P values ) by using delta delta Cq ( quantification cycle ) method , does not perform a statistical evaluation for differential expression for those genes with any number of non-detects . Therefore , in order to maximize the identification of differentially expressed genes , the non-detects in our real-time PCR results were imputed using the ImputeMissingValuesKNN module available in the GenePattern server ( The Broad Institute; http://www . broadinstitute . org ) . The imputation was limited to those genes with one non-detect among the eight samples . False discovery rates ( FDR ) were computed using p . adjust function in R statistics library ( http://cran . r-project . org/ ) . For individual mouse real-time PCR reactions of interleukin-1 beta ( IL-1β ) , we used 5’-CACAGCAGCACATCAACAAG-3’ ( forward ) and 5’- GTGCTCATGTCCTCATCCTG-3’ ( reverse ) as primer pairs . Beta actin ( Actb ) with primer pairs , 5’-ATGGAGGGGAATACAGCCC-3’ ( forward ) and 5’-TTCTTTGCAGCTCCTTCGTT-3’ ( reverse ) , was used as the endogenous control to normalize IL-1β gene expression data . We manually computed relative normalized expression using Livak’s method [49] , and P values using student’s t-test . From patient biopsies and healthy controls , total RNA was isolated using RiboPure RNA purification Kit ( Ambion ) according to manufacturer’s guidelines . cDNA synthesis was performed using the Superscript first-strand synthesis system for RT-PCR ( Invitrogen ) . The following primer sets were used: hbetaAct_F: 5′- CTCTTCCAGCCTTCCTTCCT-3′ , hbetaAct_R: 5′- AGCACTGTGTTGGCGTACAG-3′ , hIL-1β_F: 5′-GCCCTAAACAGATGAAGTGCTC-3′ , hIL-1β_R: 5′- GAACCAGCATCTTCCTCAG-3′ . The real-time PCR amplification was performed with SYBR GreenER Kit ( Invitrogen ) using an ABI Prism 7500 sequence detection system ( Applied Biosystems ) . The levels of β-actin transcription were used for normalization . We used QIAGEN’s Ingenuity Pathway Analysis ( IPA , QIAGEN Redwood City , http://www . qiagen . com/ingenuity ) to generate gene networks within the 125 significantly differentially regulated genes by uploading them along with their fold regulation and FDR values into the online application , and we chose the top 2 significant networks . Gene networks were generated based on their connectivity and the significance was measured in two ways: ( 1 ) the ratio of the number of genes from the dataset that map to the pathway divided by the total number of genes that map to the pathway; and ( 2 ) by Fischer’s exact test with P < 0 . 001 . For cryosectioning , skin tissue models were pretreated with 2M sucrose for 1h before embedding in optimum cutting temperature compound ( Sakura Finetek ) followed by freezing in liquid nitrogen and stored at -80°C . Snap frozen patient biopsies were embedded without pretreatment . 8μm cryosections were obtained using a MICROM cryostat HM 560 MV ( Carl Zeiss ) and fixed in 2% freshly prepared formaldehyde in PBS for 15 minutes at room temperature . Bacteria were visualized via conventional Gram staining . The levels of IL-1β in supernatants of infected skin tissue models , as well as plasma samples from patient 2006 and healthy controls were measured using human IL-1β Quantikine ELISA ( R&D Systems ) according to manufacturer’s guidelines . The range of detection was 0–2000 pg/ml . Multimodal distribution clusters , mean survival indices , corrected survival indices , corrected percent weight change , corrected maximum lesion area , GLM analyses , and heritability were calculated by DataDesk 6 . 3 ( Data Description , Inc . , Ithaca New York USA , www . datadesk . com ) as described previously [30 , 31] . Principal component analyses ( PCA ) and false discovery rates ( FDR ) were computed using R statistical analysis software version 3 . 2 . 1 ( http://cran . r-project . org/ ) . Other statistical analyses were performed with Prism v6 . 0d ( GraphPad Software , La Jolla California USA , www . graphpad . com ) . Survival analyses were done using the log-rank ( Mantel-Cox ) test . One-way analysis of variance ( ANOVA ) was used to evaluate differences in log-transformed bacterial loads ( skin , blood and spleens ) between the different mouse strains . We set the critical significance value ( α ) at 0 . 05 , and if the P values were less than α , we reported that the observed differences were statistically significant . | GAS bacteria are major human pathogens that are responsible for millions of infections worldwide , including severe and deadly NSTIs . Several studies have identified numerous GAS secreted virulence factors including proteases , DNases , and superantigens , which mediate several pathologic features of GAS NSTIs . However , the exact role of host genetic and/or nongenetic factors in GAS NSTIs has not been studied so far . To understand these contributions , we undertook the present study utilizing the ARI panel of BXD strains . We found that host genetic context and sex differences can modulate host-pathogen interplay and accordingly potentiate disease severity , manifestations , and outcomes . We also mapped the genetic susceptibility loci of GAS NSTIs to four mouse chromosomes , namely 2 , 6 , 7 and 18 , harboring several polymorphic genes . We believe that these findings will be helpful in uncovering further regulatory events of host-mediated GAS pathogenesis that may occur once the pathogen becomes invasive . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"dermatology",
"medicine",
"and",
"health",
"sciences",
"quantitative",
"trait",
"loci",
"biopsy",
"gene",
"regulation",
"population",
"genetics",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"animal",
"models",
"model",
"organisms",
"skin",
"infections",
"... | 2016 | Genetic Architecture of Group A Streptococcal Necrotizing Soft Tissue Infections in the Mouse |
The recent whole-genome scan for breast cancer has revealed the FGFR2 ( fibroblast growth factor receptor 2 ) gene as a locus associated with a small , but highly significant , increase in the risk of developing breast cancer . Using fine-scale genetic mapping of the region , it has been possible to narrow the causative locus to a haplotype of eight strongly linked single nucleotide polymorphisms ( SNPs ) spanning a region of 7 . 5 kilobases ( kb ) in the second intron of the FGFR2 gene . Here we describe a functional analysis to define the causative SNP , and we propose a model for a disease mechanism . Using gene expression microarray data , we observed a trend of increased FGFR2 expression in the rare homozygotes . This trend was confirmed using real-time ( RT ) PCR , with the difference between the rare and the common homozygotes yielding a Wilcox p-value of 0 . 028 . To elucidate which SNPs might be responsible for this difference , we examined protein–DNA interactions for the eight most strongly disease-associated SNPs in different breast cell lines . We identify two cis-regulatory SNPs that alter binding affinity for transcription factors Oct-1/Runx2 and C/EBPβ , and we demonstrate that both sites are occupied in vivo . In transient transfection experiments , the two SNPs can synergize giving rise to increased FGFR2 expression . We propose a model in which the Oct-1/Runx2 and C/EBPβ binding sites in the disease-associated allele are able to lead to an increase in FGFR2 gene expression , thereby increasing the propensity for tumour formation .
FGFR2 ( fibroblast growth factor receptor 2 ) plays a pivotal role both in mammary gland development and in cancer [1] . The FGFR2 gene encodes a transmembrane tyrosine kinase and can function as a mitogenic , motogenic , or angiogenic factor , depending on the cell type and/or the microenvironment . Mammary epithelial cells express FGFR2IIIb ( including alternatively spliced exon 9 ) , which binds FGF-7 and FGF-10 , which are normally expressed by surrounding mesenchymal cells . Mouse models of mammary carcinogenesis have long established the FGF signalling pathway as a major contributor to tumorigenesis [2] , and a mouse mammary tumour virus ( MMTV ) insertional mutagenesis screen for genes involved in breast cancer has identified FGFR2 and FGF10 [3] . In human breast cancer , the expression of FGFR2 has long been known to be elevated in estrogen receptor ( ER ) –positive tumours [4] , which has been confirmed by data analysis performed with the ONCOMINE 3 . 0 array database [5 , 6] . Likewise both FGF-7 and FGF-10 have been found to be expressed in a proportion of breast cancers [7 , 8] . Functional studies in cell lines have implicated FGFR2 as playing a role in tumourigenesis , with an alternative splicing in the C-terminal domain of FGFR2 giving rise to a more strongly transforming isoform [9] . However , as yet , nothing is known about the mechanism by which FGFR2 acts as a risk factor in predisposition to breast cancer . We examined the functional implication of genetic variation in the FGFR2 haplotype associated with susceptibility to breast cancer and we demonstrate increased gene expression for the risk allele .
Two independent studies have identified FGFR2 as risk factor in breast cancer [10 , 11] . We have shown that in Europeans , the minor disease-predisposing allele of FGFR2 is inherited as a haplotype of eight single nucleotide polymorphisms ( SNPs ) covering a region of 7 . 5 kb within intron 2 of the gene [10] ( Figure 1 ) , in a haplotype block with no linkage disequilibrium with the coding region of the gene . Microarray gene expression analysis on the Nottingham City Hospital cohort , using both the Agilent [12] and the Illumina [13] platforms , indicated that FGFR2 is expressed at higher levels by tumours that are homozygous for the minor alleles than by those with the common alleles ( Wilcox p < 0 . 05 ) . Analysed tumours were all diploid for this region based on array-comparative genome hybridization data [14] . This correlation was independent of either ER expression or p53 mutation status of the cells . Quantitative TaqMan PCR analysis confirmed a significant increase in FGFR2 expression in rare homozygotes , as compared to common homozygotes ( Wilcox p = 0 . 028 ) ( Figure 2 ) . We also examined expression of the FGFR2 ligands FGF-7 , FGF-10 , and FGF-22 , which are usually produced by the surrounding stroma , in 45 normal breast samples as well as the microarray data on tumours , but we found no correlation with genotype . Furthermore , FGFR2 displays a very complex splicing pattern with the most commonly expressed variants of the N terminus of the gene either including exons 1 , 2 , and 3 or including exons 1 and 2 , but lacking exon 3 . Again , no correlation was observed between genotype and the presence or absence of exon 3 . Thus , the risk genotype correlates with FGFR2 expression itself , rather than affecting its function through receptor-ligand interactions . This correlation suggests that the functional SNPs map to a regulatory region within the gene , possibly by altering one or more transcription factor binding sites . Interactions between proteins from nuclear extracts and DNA were examined for the eight most strongly disease-associated alleles ( Figure 1 ) . Two of these candidate functional SNPs showed distinct binding patterns in electrophoretic mobility shift assays ( EMSA ) . The common allele of rs7895676 ( FGFR2–33 ) formed strong protein–DNA complexes with nuclear extracts from the breast carcinoma cell lines HCC1954 ( Figure 3A ) and PMC42 and from HeLa cells ( unpublished data ) , whereas no binding was detected on the minor allele . Competition studies and supershift experiments identify the bound protein as C/EBPβ ( Figure 3A ) . We note that the FGFR2–33 sequence has considerable homology to the C/EBPβ binding site from the interleukin 6 ( IL-6 ) promoter [15] ( Figure 3C ) . The heterogeneity of the observed protein–DNA complexes is most likely due to the presence of multiple C/EBPβ isoforms . For rs2981578 ( FGFR2–13 ) , both alleles give rise to a strong protein–DNA complex in HCC1954 cell extracts . However , a second more slowly migrating complex was only seen on the rarer genotype ( Figure 3B ) . Interestingly , both alleles are able to compete for both bands , suggesting that the formation of the upper complex depends on the presence of the lower complex . Inspection of the FGFR2 DNA indicated the presence of a perfect octamer binding site immediately adjacent to the SNP , while the SNP itself lay within a sequence with homology to Runx binding sites ( Figure 3C ) . Competition studies and incubation with specific antisera shows that both alleles bind Oct-1 , while only the minor allele binds Oct-1 and Runx2 in HCC1954 nuclear extracts ( Figure 3B ) , as well as in PMC42 cells ( Figure S1 ) . To establish whether or not these sites were occupied in vivo , we carried out chromatin immunoprecipitation ( ChIP ) experiments using the ER+ breast cancer cell lines HCC70 and T47D , which are homozygous for the minor and the common FGFR2 alleles , respectively . In addition , we confirmed that these cell lines were diploid for the FGFR2 locus and only expressed the epithelial-specific isoform FGFR2IIIb [16] . The ChIP analysis was carried out on homozygous cell lines , because the SNP overlapping the C/EBPβ site lies in a repetitive region for which the different alleles could not be distinguished reliably by TaqMan PCR . A representative experiment is shown in Figure 3D . After Runx2-precipitation , the FGFR2–13 site is enriched 2-fold for the minor versus the common allele , confirming the EMSA results . Western blotting indicated that Runx2 is more abundant in T47D cells , thus confirming that differential ChIP in the two cell lines is due to the presence of the SNP . Oct-1 precipitation did not yield enrichment of FGFR2–13 for either allele . The Oct-1 epitope may either be sequestered within a higher-order complex or the antisera used do not work efficiently in a ChIP assay . On the FGFR2–33 site , we observed a 1 . 7-fold enrichment of C/EBPβ binding on the common allele . In addition , we observe that C/EBPβ can also bind to the minor allele , although less efficiently . Both cell lines contain comparable amounts of C/EBPβ as judged by Western blotting ( unpublished data ) . In conclusion , both the C/EBPβ and the Runx2 binding sites are occupied in vivo . To test whether differential protein binding could alter the ability of the susceptibility alleles to activate transcription , we multimerised oligonucleotides overlapping both the Oct-1/Runx2 and the C/EBPβ binding sites , cloned these in both orientations upstream of the luciferase reporter gene in pGL3Enh ( Figure 4A ) , and assayed them in three breast cancer cell lines ( PMC42 , HCC70 , and T47D ) . Transfections were carried out in triplicate and repeated at least twice for each cell line . A representative transfection into HCC70 cells is shown in Figure 4B ( see Figure S2 for PMC42 and T47D ) . In all three cell lines tested , the minor allele at the Oct-1/Runx2 site stimulated transcription 2- to 5-fold over the common allele , independent of orientation , with the average being just above a 3-fold increase ( p < 0 . 01 ) . In contrast , the minor and common alleles of the multimerised C/EBPβ binding site did not show a consistent pattern of activation relative to each other . It varied with the cell lines and the orientation in which constructs were tested . Nevertheless , relative to the parental vector , the common allele always showed transcriptional activation . Compared to the common allele , the minor allele was either not significantly different or gave rise to a smaller degree of activation . However , in the latter case , the rare allele still activated transcription significantly above the enhancer-only construct ( p < 0 . 01 ) . Presumably this reflects the fact that the minor allele of FGFR2–33 still binds C/EBPβ above background levels in vivo ( Figure 3D ) . By comparing the two different sites , we found that for Oct-1/Runx2 the minor allele was more active , while for C/EBPβ , the common site yielded higher levels of transcription in the majority of experiments . Hence their effects were opposing . We therefore assayed a synthetic construct consisting of single sites for C/EBPβ , Oct-1 , and Runx2 . In this arrangement , the effect of Oct-1/Runx2 clearly predominates , with the minor allele expressed at higher levels , reflecting the situation at the endogenous locus .
The data presented here lead us to conclude that the Oct-1/Runx2 binding site is the dominant determinant of differential expression between the common and minor haplotypes of FGFR2 . Although Runx2 is a master regulator of osteoclast-specific transcription , Runx2 also plays an important role in mouse mammary gland–specific gene expression [17] , where Runx2 activity is dependent on Oct-1 [18] . It is intriguing to note that in bone cells , overexpression of constitutively active FGFR2 leads to increased levels of Runx2 mRNA [19] . FGFR2 in turn is responsive to Runx2 in osteoclasts via the OSE2 ( osteoclast specific element 2 ) in its promoter [20] . The description here of a Runx2 site in the FGFR2 gene that is occupied in breast cancer cells , suggests that in the presence of the minor genotype , a similar positive feedback loop could also be established in breast cells . The role of the C/EBPβ binding site on FGFR2 expression has been harder to define . The common allele binds C/EBPβ more tightly and activates transcription more strongly in most cases . Yet in a composite construct the activity of the Oct-1/Runx2 site dominates . This may be because C/EBPβ can directly bind to and synergize with Runx2 [21] . Thus , on the minor genotype , Oct-1 and Runx2 are present and able to synergize with the C/EBPβ bound ( as suggested from the ChIP experiments ) , giving rise to higher levels of transcriptional activation . This is supported by the finding that a single copy of the C/EBPβ/Oct-1/Runx2 site gives rise to higher levels of activation than a concatemerized Oct-1/Runx2 site with six potential interaction sites ( Figure 4A ) . A potential role for C/EBPβ in tumour etiology is supported by the observation that C/EBPβ is highly overexpressed in malignant human breast cells [22] . In conclusion , our evidence supports Oct-1/Runx2 as the probable primary determinant of activity , with C/EBPβ contributing to the risk haplotype . The increased risk in breast cancer conferred by the FGFR2 allele is predominant for ER+ breast tumours , while there is no significant increase in risk for ER– tumours . Genome-wide analysis of ER binding sites has revealed three potential ER binding sites within the FGFR2 gene [23] , and ER and Oct-1/Runx2 may cooperate to increase gene expression . This is consistent with findings that Oct and ER sites often cluster [23] . The risk conferred by the disease-associated genotype may also depend on the signalling potential of FGFR2 in ER+ cells . FGF-7 is over-expressed only in breast tumours that are ER+ [8] . Elevated levels of FGFR2 may then contribute to the establishment of an autocrine signalling loop , reducing the cell's propensity to undergo apoptosis [24] . Alternatively , paracrine signalling by mesenchymally or luminally derived FGF-7 or -10 on cells overexpressing FGFR2 may also drive cell proliferation . To our knowledge , this is the first functional study on the risk loci recently identified for breast cancer . Our study demonstrates that SNPs identified by whole-genome scans can be used a valid starting points for studying the underlying biology of cancer . SNPs identified in other whole-genome scans for the genetic basis of complex diseases also primarily map in intronic or intergenic regions . Our observation that an identified SNP regulates the expression of the risk allele is therefore likely to be a common theme . Breast cancer is one of the most common cancers in the developed world . The FGFR2 minor allele carries only a small increase in risk and acts as part of a spectrum of risk factors . However , it has a high minor allele frequency ( 0 . 4 ) , and FGFR2 is therefore likely to contribute to the incidence of breast cancer in many individuals .
DNA from the 170 tumour samples was genotyped using a fluorescent 5′ exonuclease assay ( TaqMan ) and the ABI PRISM 7900 Sequence Detection Sequence ( PE Biosystems ) in 384-well format . Duplicate samples were included to assess concordance and quality of genotyping . The genotyping assay was designed for rs2981582 , which tags the whole haplotype block associated with the disease [10] . Analysis was performed on total RNA from breast tumour cases . cDNA was prepared with the TaqMan Reverse Transcription Reagents kit ( Applied Biosystems ) using random hexamers , according to the manufacturer's instructions . Expression levels were determined using a TaqMan Gene Expression Assay ( Hs00240796_m1 , Applied Biosystems ) and normalized to four different housekeeping genes . To assess whether there were significant statistical differences between the expression levels across the genotype groups we used a Wilcoxon test , fitted using the R statistical framework . Elsewhere , Student's t-tests were carried out using Microsoft Excel . Breast cancer cell lines HCC1954 , HCC70 , T47D , and PMC42 were cultured in RPMI supplemented with 10% foetal calf serum and penicillin/streptomycin under standard conditions . These cell lines have been characterised extensively , and karyotypes are available at the Cancer Genomics Program of the University of Cambridge ( http://www . path . cam . ac . uk/~pawefish ) . Small-scale nuclear extracts and bandshifts were carried out as previously described [25] , except that Complete Protease Inhibitors ( Roche ) were used . In supershift experiments , polyclonal antisera against Oct-1 ( sc-232x ) , Runx2 ( sc-10758x ) , and C/EBPβ ( sc-150x ) were obtained from Santa Cruz Biotechnology , Inc and up to 8 μl were added per reaction , unless otherwise stated . Oligonucleotides ( Table S1 ) were annealed to complementary strands , and the resulting BamHI overhangs filled in with Klenow enzyme , using radiolabelled [α32P]dCTP ( GE Healthcare , UK ) . Primers were designed using Primer Express ( Applied Biosystems ) and Lasergene ( DNA Star ) to amplify regions of up to 100 bp comprising the SNPs of interest , plus one negative control ( region of the genome not suspected to bind any of the transcription factors of interest ) ( Table S1 ) . PCR amplification was carried out with Power SYBR Green Mastermix ( Applied Biosystems ) , using 2 μl of precipitated and purified DNA as described [23] . The antisera were as in the EMSAs , except for C/EBPβ , which was a polyclonal serum from Abcam , UK . The pGL3-Enhancer vector ( Promega ) was linearized with BglII and re-circularised in the presence of annealed oligonucleotides ( Table S1 ) . All constructs were verified by sequencing . DNA was prepared using Qiagen kits and transfected into tumour cell lines cultured in 24-well plates . Per well , 500 ng of reporter and 100 ng CMV-β-galactosidase plasmid were tranfected using 2 μl of Fugene 6 ( Roche ) , harvested 36–48 h later and extracts prepared using 100 μl Promega lysis buffer . Luciferase and β-galactosidase activity in 25 μl was measured using Promega reagents . Results are given as ratios of luciferase over β-galactosidase activity . | Recently , a number of whole-genome association studies have identified genes that predispose individuals to common diseases such as cancer . The challenge now is to understand how the identified risk loci contribute to disease , since the majority of these loci are located within introns ( which are discarded after transcription ) and intergenic regions , and therefore do not change the coding region of nearby genes . This manuscript describes how two single–base pair changes in intron 2 of the FGFR2 ( fibroblast growth factor receptor 2 ) gene , “the top hit” of the breast cancer susceptibility study , exert their function . We find that the changes alter the binding of two transcription factors and cause an increase in FGFR2 gene expression , thus providing a molecular explanation for the risk phenotype . This is the first functional study , to our knowledge , of the risk loci identified for breast cancer in a whole-genome scan and demonstrates that these studies can be used as valid starting points for studying the underlying biology of cancer . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics"
] | 2008 | Allele-Specific Up-Regulation of FGFR2 Increases Susceptibility to Breast Cancer |
Chromosome 3p21–22 harbors two clusters of chemokine receptor genes , several of which serve as major or minor coreceptors of HIV-1 . Although the genetic association of CCR5 and CCR2 variants with HIV-1 pathogenesis is well known , the role of variation in other nearby chemokine receptor genes remain unresolved . We genotyped exonic single nucleotide polymorphisms ( SNPs ) in chemokine receptor genes: CCR3 , CCRL2 , and CXCR6 ( at 3p21 ) and CCR8 and CX3CR1 ( at 3p22 ) , the majority of which were non-synonymous . The individual SNPs were tested for their effects on disease progression and outcomes in five treatment-naïve HIV-1/AIDS natural history cohorts . In addition to the known CCR5 and CCR2 associations , significant associations were identified for CCR3 , CCR8 , and CCRL2 on progression to AIDS . A multivariate survival analysis pointed to a previously undetected association of a non-conservative amino acid change F167Y in CCRL2 with AIDS progression: 167F is associated with accelerated progression to AIDS ( RH = 1 . 90 , P = 0 . 002 , corrected ) . Further analysis indicated that CCRL2-167F was specifically associated with more rapid development of pneumocystis pneumonia ( PCP ) ( RH = 2 . 84 , 95% CI 1 . 28–6 . 31 ) among four major AIDS–defining conditions . Considering the newly defined role of CCRL2 in lung dendritic cell trafficking , this atypical chemokine receptor may affect PCP through immune regulation and inducing inflammation .
Functional variation in the human leukocyte antigen ( HLA ) class I genes and in chemokine receptors affects HIV susceptibility , viral load , and rates of disease progression [1]–[4] . Recent genome-wide association studies ( GWAS ) performed in HIV-1 cohorts have shown that the HLA region and the chemokine receptor CCR5 gene have major roles in control of HIV-1 replication and disease progression—together they explain approximately 20% of genetic variability [3] , [5]–[8] ( reviewed in [4] , [5] ) . These findings from GWAS highlighted the leading role of chemokine receptors among non-HLA genes in HIV-1 pathogenesis and prompted us to assess the role of other chemokine receptor genes on HIV disease using a gene-centric approach to identify common or rare functional variants in the region . The chemokine receptor cluster on chromosome 3 contains at least 12 genes including CCR5 , the primary HIV-1 co-receptor [9]–[11] . Multiple genetic variants in chemokine receptors and chemokines have been identified as modifiers of HIV-1 infection or disease progression [12] , [13] , including CCR5-Δ32 ( a 32-bp deletion introduces a premature stop codon ) [14] and CCR5 promoter variants [15]–[17] and variants in the CCR5 ligand gene CCL5 [18] , [19] . The homozygous CCR5 Δ32/Δ32 genotype and complex heterozygotes with other rare amino acid mutations confers near complete resistance to HIV infection [12] , [14] , [20]–[22] . Individuals homozygous for a haplotype known as CCR5-P1 [15] or haplogroup HHE [23] , a multisite allele of the CCR5 promoter region , progress to AIDS more rapidly than those with other CCR5 promoter haplotypes [15]–[17] , [23]–[25] . CCR2 and CXCR6 are minor HIV-1 coreceptors used by a limited number of HIV-1 strains as an entry coreceptor [26] , [27] . CCR2-V64I has been associated with delayed progression [12] , [25] , [28] , [29] . Variants in CXCR6 were also associated with disease modification [30] , [31] . The chromosome 3 chemokine receptor cluster extends from 3p21 to 3p24 , with eight receptors occurring in an 520 kb region of 3p21 ( Figure 1 ) [32] . The cluster contains genes for several receptors , CCR3 , CCR8 , CX3CR1 , and CXCR6 that have been shown to bind HIV env or to support varying levels of in vitro replication of HIV-1 , HIV-2 or simian immunodeficiency virus ( SIV ) [26] , [33]–[36] ( reviewed by [11] , [37] ) . The role played by minor coreceptors in HIV-1 pathogenesis is not clear , but studies have suggested that a broad spectrum of coreceptor usage may be correlated with rapid CD4+ cell depletion and AIDS progression [11] , [38] , [39] . Primary isolates of HIV-1 have been shown to use a wide spectrum of various chemokine receptors as HIV coreceptors [40] . HIV-1 isolates from a CCR5-Δ32 heterozygous or homozygous individual can use various minor coreceptors such as CCR3 , CCR2B , CCR8 , CX3CR1 for cell entry [41] , [42]; amino acid mutations in the V3 loop of HIV-1 are responsible for utilization of multiple coreceptors [40] . Considering the high mutation rate and sequence heterogeneity of HIV-1 , particularly within the env gene , it is plausible that a spectrum of receptors is used in vivo during the course of HIV infection and that genetic variants in the coreceptors may affect usage or binding efficiency by HIV-1 . Furthermore , as CCR5 and CXCR4 antagonists blocking these major co-receptors are used therapeutically [37] , the potential of HIV-1 to evolve to use other minor coreceptors as alternative cell entry points is expected to increase . Therefore , determining whether HIV-1 minor coreceptor genes , in addition to CCR5 and CXCR4 play a role in HIV pathogenesis is a timely topic . In this study , we evaluate the impact of chemokine coreceptors ( CCR ) on HIV/AIDS using a candidate-gene based population association analysis in five treatment-naive HIV-1 natural history cohorts . Genotypes of exonic polymorphisms in chemokine coreceptor genes CCR3 , CCRL2 and CXCR6 on Chromosome 3p21 and in CCR8 and CX3CR1 on 3p22 were tested for their genetic influence on AIDS progression . CCR3 , CCR8 and CXCR6 were chosen as they are HIV-1 minor coreceptors [26] , [33]–[36] ( reviewed by [11] , [37] ) . CCRL2 was selected as a candidate gene because of its homology with CCR5 ( 45% ) —the most of any of the chemokine receptors genes—because of its proximity to CCR5 , and because it is an atypical receptor without signal transduction , similar to DARC . Our results suggest that genetic variation in CCR3 , CCR8 and CCRL2 may contribute additional genetic regulation of HIV-1 disease in addition to that conferred by the major HIV-1 coreceptor gene CCR5 .
We resequenced selected chemokine receptor genes in the chromosome 3p21–22 region in 72 African Americans and 72 European Americans with three extreme phenotypes ( resistance to HIV infection , very rapid or slow progression to AIDS ) to assess the extent of exonic variation and to identify rare variants . We observed a total of 6 exonic variants , 5 of which were nonsynonymous , in CCR3 , CCR8 , CXCR6 , and CCRL2 ( Figure 1 , Table 1 ) . We did not resequence CCR5 and CCR2 since these receptor genes had been previously sequenced in HIV patients [28] , [43] , nor did we resequence CX3CR1 . We previously showed that CX3CR1-V249I and -T280M had no effect on AIDS progression in our group of seroconverter subjects [44] , and therefore did not include them in this analysis . No noticeable differences in SNP frequency among three extreme phenotype groups were observed ( data not shown ) . These SNPs were then genotyped in 5 HIV-1 natural cohorts comprising 2594 European Americans ( EA ) . All SNPs were in Hardy-Weinberg Equilibrium ( P>0 . 05 ) . CXCR6-E3K and CCR3-P39L were rare ( <1% ) in EA and were excluded from the analysis; other SNPs were common ( >5% ) ( Table 1 ) . The SNPs considered fell into two distinct haplotype blocks ( Figure 1 ) . The larger block in 3p21 comprises CCRL2-F167Y , CCRL2-I243V , CCR5-+/Δ32 , the CCR5-promoter SNP CCR5-2459A ( rs1799864 , previously shown to affect CCR5 expression levels and modify AIDS progression [15] , [16] ) , CCR2-V64I , CCR3-P39L and CCR3-255T/C ( Y17Y ) , spanning 77 Kb , and CXCR6-E3K , 318 Kb teleomeric to CCR3 . The second block in 3p22 , ∼10 Mb from the first block , consists of CX3CR1-V249I and -T280M and CCR8-A27G ( rs2853699 ) , which are separated by ∼160 Kb . There is moderate to substantial LD within the closely spaced CCR3-CCR2-CCR5-CCRL2 group ( with pairwise D′ 0 . 47–1 , Figure S1 ) . In the 3p22 block , low levels of LD were observed between CCR8 and CX3CR1 SNPs ( D′ ∼0 . 40 ) . The LD level between 3p21 and 3p22 blocks is minimal . The identification of independent additional genetic factors in this region , particularly for 3p21 , is complicated by a moderate to high level of LD ( Figure S1 ) . To detect genes or markers additional to a primary predisposing variant ( s ) in a genetic region of high LD , stratification analyses or using a restricted dataset are frequently employed [45] . Stepwise regression is also a choice for assessing the relative importance of different variants in the linked region [46] . These approaches are feasible as the SNPs in this region have a low to moderate correlation ( for r2 , see Figure 2 ) . SNPs with allele frequencies of at least 5% were analyzed by Kaplan-Meier survival curve analysis and Cox proportional hazards model . We present the relative hazards ( RH ) of the individual variants on survival to AIDS from analysis of 670 EA seroconverters using multivariable Cox regression models in Table 2 . Four of the five SNPs in the 3p21 block revealed new significant associations with delayed time progressing to AIDS , after conditioning on HLA alleles and CCR5-Δ32 ( Table 2 ) . Significant associations with differential progression to clinical AIDS were observed for CCR3 -255C ( RH = 0 . 62 , P = 0 . 009 , Figure 2A ) , for CCRL2-243V ( RH = 0 . 66 , P = 0 . 03 , Figure 2B ) , and for CCRL2-167F ( RH = 1 . 89 , P = 0 . 0003 , Figure 3A ) and for CCR8-27G ( RH = 1 . 44 , P = 0 . 004 ) ( Table 2 ) . Adjusting for potential population stratification had little affect on the significance of the associations ( Table 2 ) . Correcting for 7 chemokine receptor variants tested using a Bonferroni step-down ( Holm ) correction [47] , the associations for CCRL2-167Y , CCRL2-243V , and CCR3-255C remained significant ( P = 0 . 002 , 0 . 03 , and 0 . 016 , respectively ) , in addition to CCR5-Δ32 and CCR5-2459 , while CCR8-27G became non-significant ( Table 2 ) . Using stepwise selection determined by the Akaike information criteria ( AIC ) , we built an optimal prediction model with the genetic variants that were associated with AIDS progression in this or previous studies ( Table 3 ) . From Table 3 , we can see the minimum AIC is achieved at the ninth step with nine covariates . Therefore , the best predictive model accounting for the rate of progression to AIDS ( 1987 CDC definition ) includes the newly identified variants CCRL2-167F and -243V , CCR8-27G , as well as previously known variants: the two locus CCR2-V64I-CCR5-Δ32 composite genotype , CCR5-2459 , HLA class 1 homozygosity , HLA-B*35Px , HLA-B*57 and HLA B*27 ( without covariates , AIC:1796 . 32 , with 9 covariates , AIC:1737 . 28 ) . The significance level and RH values in this model represent that of each variant after considering all other linked and unlinked variants ( Table 3 ) . The rough shrinkage estimate is 0 . 88 ( 104/140 ) , indicating that the model is fairly reliable without further shrinkage or data reduction treatment [48] . We calculated that R2 = 0 . 11 for the final fitted model , indicating that 11% of the variability in AIDS progression can be explained with the variants in the model . The strongest new association was for CCRL2-Y167F with progression to clinical AIDS . The common 167F allele ( allele frequency 67% in EA ) had a significant dominant detrimental effect on progression to AIDS-defining conditions ( CDC 1987 case definition ) ( RH = 1 . 89 , P = 0 . 0003 , Wald test; P = 0 . 0001 , Likelihood ratio test; n = 670 seroconverters ) in the Cox proportional hazards model ( Table 2 and Table 3 , Figure 3A ) . When the analysis was restricted to individuals not carrying CCR5-Δ32 ( n = 539 ) , the association of CCRL2-167F with AIDS remained significant ( RH = 1 . 69 , 95% CI 1 . 20–2 . 38 ) , indicating that the CCRL2-167F association is independent of CCR5-Δ32 and not due to its linkage disequilibrium with the latter . In an explanatory analysis , we tested whether CCRL2- Y167F affects AIDS progression through a specific AIDS-defining illness . A Cox model analysis in the seroconverter group showed significantly faster progression to PCP for carriers of 167F ( RH = 2 . 84 , 95% CI 1 . 28–6 . 31 , P = 0 . 007; Figure 3B ) , but no differential effect on Kaposi's sarcoma , microbacterial avium infection , cytomegalovirus , or lymphoma ( P>0 . 05 , data not shown ) . To assess whether the CCRL2-167F association with PCP was due to LD ( D′ = 0 . 76 , r2 = 0 . 47 ) with CCR5-2459 that associated with AIDS progression [16] , we restricted the analysis in a group of SC ( n = 365 ) that do not carry the homozygous CCR5-2459 genotype; the CCRL2-167F association with PCP remained significant ( OR = 2 . 42 , 95% CI 1 . 09–5 . 36 ) , in support of its independent role .
We investigated the impact of genetic variation of 3p21–22 chemokine receptor genes on HIV/AIDS in this population-based association study . We identified a total of 6 exonic variants in CCR3 , CCR8 , CXCR6 and CCRL2 in 144 samples with extreme phenotypes . Through Cox regression survival analysis of the variants in HIV-1 natural cohorts , we identified three polymorphisms ( CCRL2-Y167F , CCR3-255C and CCR8-27G ) as having previously unidentified correlation with AIDS progression that appear to confer additional effect beyond the well-studied AIDS-modifying polymorphisms CCR5-Δ32 , CCR5-P1/CCR5-2459A , and CCR2-64I . CCRL2-167F was associated with strong accelerated progression to AIDS , resulted almost entirely from rapid development of the AIDS-defining disease PCP . The coverage of SNPs in the 3p21–22 region is limited in the GWAS SNP chips . Several exonic SNPs ( CCRL2-167F , -243V , CCR8-27G , CCR5-2459A , CCR5- Δ32 ) genotyped in this study were not included in the SNP chips that have been used in GWAS for HIV-1 [3] , [5]–[8] , highlighting the ongoing need of candidate gene analysis in the GWAS era . The newly observed effects of chemokine coreceptor genes thus await further replication . It must be cautioned that the combination of the presence of multiple AIDS associations in this chemokine receptor complex , and the LD between the receptor genes , makes determining the true source of the associations difficult . Overall , however , the association analysis points to receptor gene variant associations beyond those that can be attributed to the known AIDS affecting receptor gene polymorphisms . The association of additional chemokine receptors , beyond the primary CCR5 and CXCR4 , with AIDS progression is plausible as several of these have been shown to bind to HIV in varying degrees [26] , [35] , [36] , [39]–[42] , [53] . The complexity of associations in this region makes it essential to identify a functional effect of the genetic variants on disease , before concluding that the associations are real . It should be noted that we may have not detected all existing SNPs in the region and was also underpowered in detecting rare SNPs ( 1% ) . We found that another CCRL2 exonic SNP in the public domain ( 168M , rs6441977 ) was not associated with progression to AIDS-87 ( RH = 0 . 84 , 95% 0 . 50–1 . 39 ) in our seroconverters . We performed haplotype analysis for the three exonic SNPs in CCRL2 and found that only the haplotypes bearing rs3204849 A ( CCRL2 Y167F ) and rs3204850 ( V243I ) were associated with AIDS; the haplotype bearing rs6441977A ( V168M ) had no effect . No additional information was gained by performing a haplotype analysis compared to single SNP analysis . With these caveats we argue that the association of the CCRL2-167F variant is worthy of interest . First , within the noted limits of the association analysis , the association is quite strong ( RH = 1 . 9 , P = 0 . 002 , corrected ) , remains significant with CCR5-Δ32 taken into account , and in an automatic selection analysis with all known factors taken into account . Second , although there is no direct demonstration of function for this polymorphism , and the F to Y substitution is generally a conservative one , several lines of phylogenetic , chemical modeling , and indirect experimental data suggest that CCRL2-167Y significantly alters the properties of this receptor . Protein structure modeling data suggest that the risk-associated F to Y substitution could change the boundary of the transmembrane domain , and introduce a hydrogen bond . Of particular interest is the alignment data of the conservation of this residue among CC chemokine receptors . Strikingly , phenylalanine , the ancestral alternate variant to the AIDS risk tyrosine variant , or another nonpolar amino acid , occurs at this position in all receptors known to be functional [54]; a polar amino acid only occurs in the two cases of the nonsignaling DARC receptor . Further , substitution of tyrosine for phenylalanine at this position in the functional receptor CCR3 reduced migration of HEK293 cells in response to eotaxin threefold . We emphasize that all of these tests are in silico or indirect , and direct test of the effect of the F to Y substitution on the function of CCRL2 remains to be done; new knowledge of the functions and ligands of CCRL2 should make this more straightforward . The association of CCRL2 with AIDS and PCP is unique as CCRL2 has not been shown to serve as a coreceptor for HIV-1 . Our chemotaxis assay experiments excluded 11 common chemokines as ligands of CCRL2 . The association does not appear to be a general function of increased HIV susceptibility , but instead specifically attributable to an increase in PCP among individuals carrying this receptor variant . PCP , caused by Pneumocystis jirovecii , is the most common opportunistic infection in untreated HIV-1-infected immunosuppressed persons . PCP is mediated by marked inflammatory responses in lung involving macrophages and chemokines and cytokines [55] . The association of CCL2 with PCP might be attributable to two aspects of CCRL2: immune regulation or direct interaction with HIV . CCRL2 may affect PCP through its immune regulating role at local inflammation sites , possibly by concentrating and presenting chemokines [51] , [52] , [56] . CCRL2 was rapidly upregulated in murine lung macrophages following inflammation induction [57] and deficiency of CCRL2 impaired lung dendritic cell migration [58] . Alternatively , CCRL2 may influence HIV coreceptors entry through interacting or sequestering with CCL5 or anti-HIV chemokines [56] , or may serve as a coreceptor for some strains of HIV-1 . The mechanism of CCRL2-F167Y effect on PCP remains to be explored . In summary , this comprehensive study of the chromosome 3 chemokine receptor cluster region identified multiple genetic variants that associated with HIV disease . The strongest new association appears to result from an increased susceptibility to PCP , rather than from a specific effect on HIV . Added to the existing knowledge of the effect of the chromosome 3 chemokine receptors on HIV disease , which has been already been exploited therapeutically , our results affirm this gene complex as a fertile ground for further research , both for HIV and potentially for a broad range of additional diseases .
Institutional review boards ( IRB ) at National Cancer Institute , National Institutes of Health and participating institutes approved the study protocols . Written informed consent was obtained from all study participants and/or their legal guardians . The study group includes 674 HIV-1 seroconverter European Americans , 669 seronegatives enrolled in the following natural history HIV-1 cohort studies: Multicenter AIDS Cohort Study ( MACS ) [59] , the San Francisco City Clinic Cohort Study ( SFCCC ) [60] , AIDS Link to the Intravenous Experience ( ALIVE ) [61] , Hemophilia Growth and Development Study ( HGDS ) [62] , and the Multicenter Hemophilia Cohort Study ( MHCS ) [63] . Seroconversion date was estimated as the midpoint between the last seronegative and the first seropositive HIV-1 antibody test date ( mean interval 0 . 79 years , range 0 . 07–3 . 0 years ) . The censoring date was the earliest of the date of the last recorded visit , or July 31 , 1997 for the ALIVE cohort , or December 31 , 1995 for all other cohorts , to avoid the confounding effect of highly active anti-retroviral therapy ( HAART ) . A later censoring date was used for ALIVE cohort because few ALIVE participants received HAART prior to July 31 , 1997 [64] . A panel of 72 EA and 72 AA samples representing extreme phenotypes for infection and progression ( rapid progression , long term non-progression , and infection resistance ) were used for SNP identification . The 5′ and 3′ untranslated ( UTR ) and coding regions of the CCR3 , CCR8 , CCRL2 , and CXCR6 genes were PCR amplified by overlapping primer sets ( Table S2 ) . PCR products were resequenced by BigDye terminator ( Applied Biosystems , Foster City , CA ) . We did not sequence or genotyping SNPs in CCR1 , CCXCR1 , CCR9 and CCR4 as they are not recognized as HIV-1 coreceptors [26] , [33]–[36] ( reviewed by [11] , [37] ) . Genotyping was done using PCR-restriction fragment length polymorphism ( RFLP ) or TaqMan assays . PCR primer sequences , TaqMan probes and primers , PCR conditions , and restriction enzymes used to genotype each variant are listed in Table S1 . Briefly , PCR was carried out with 35 cycles of denaturing at 94°C for 30 s , annealing at 54–60°C for 30 s and extension at 72°C for 45 s . TaqMan assays were performed according to the manufacturer's manual ( Applied Biosystems , Foster City , CA ) . The CX3CR1 variants V249I and T280M were typed as previously reported [65] . Kaplan-Meier survival statistics and the Cox proportional hazards ( PH ) model ( Cox PH model ) were used to assess the effects of genetic variants on the time of progression from HIV-1 infection to AIDS ( 1987 CDC definition ) [66] , using PROC PHREG and LIFETEST of SAS version 9 . 13 ( SAS , Cary , North Carolina ) . For SNPs that showed significant association with AIDS development , explanatory analyses were performed for their specific impact on the AIDS-defining diseases Pneumocystis pneumonia ( PCP ) , Kaposi's sarcoma , microbacterial avium infection , and cytomegalovirus . The relative hazard ( RH ) and significance of associations were determined using a Cox PH model without or with adjustment for confounding genetic factors not on chromosome 3: for EA HLA-B*27 and HLA-B*57 , HLA-B*35Px , and HLA Class I homozygosity [1] , [67] , [68]; for AA CCRL5-In1 . 1 , HLA-B*57 and HLA Class I homozygosity [1] , [19] , [67] . CCR2-64I , CCR5-Δ32 and CCR5-2459 were also included as covariates in the adjusted multivariable regression analysis . CCR5 promoter haplotypes ( P1 , P2 , and P4 ) are tagged by SNPs CCR5-2459 ( rs1799987 ) and rs2734648 [43] . A visual inspection of the data with Kaplan-Meier survival curves was performed to determine the genetic models to be used in the Cox PH regression model . A dominant genetic model was tested for all genetic factors in this study , except for CCR5-2459 ( recessive ) [15] , [16] . Participants were stratified by sex and by age at seroconversion: 0–20 , >20–40 , and over 40 years . To determine the best explanatory set of genetic variants , while minimizing the number of comparisons in model selection , We used StepAIC procedure to build Cox proportional hazards models for the AIDS-1987 phenotype based on stepwise regression , Akaike information criteria ( AIC ) , and the best subset selection [69] . Here we used PROC PHREG in SAS software with SLENTRY = 0 . 99 and SLSTAY = 0 . 995 ( values chosen close to 1 to generate a sequence of models from null models to full model ordered by AIC ) [48] . Model uncertainty caused by including large number of variables can be estimated by shrinkage ( LR-p ) /LR , where LR denotes the likelihood ratio χ2 and p denotes the number of the predictors in the final model . A shrinkage below 0 . 85 raises concern of overfitting [48] . It is recommended that no more than m ( the number of uncensored event ) /10 predictor degree of freedom p ( number of parameters ) should be examined to fit a multiple regression model . As in our sample , there were 194 events without censoring , we expect that fitting with <19 variables would be appropriate [48] . To control for potential population stratification , we adjusted the regression analysis with eigenvector values [70] . Eigenvector values were obtained by performing principal component analysis of 700 , 022 SNPs from a previous GWAS study carried out in the same samples [71] . In the Cox regression analysis , eigenvector values for the top two principle components were included as covariates . The genomic inflation factor in this seroconverter population showed minimal systematic overall bias due to population structure in regards to disease progression phenotype as it was quite close to 1 . 0 ( λ = 1 . 01 ) ( expected under no population stratification ) [71] . We further corrected for multiple comparisons by counting 7 chemokine receptor variants tested using a Bonferroni step-down ( Holm ) correction method [47] , as implemented in the MULTITEST procedure in SAS . The “generalized” R2 statistic in Cox model is based on the likelihood-ratio statistic ( LRT ) for testing the global null hypothesis [72] . The formula is given as: R2 = 1−e− ( LRT/n ) , where LRT = −2logL ( 0 ) −[−2logL ( p ) ] , n is the total sample size , logL ( 0 ) is the log-likelihood for a null model with no covariates , and logL ( p ) is the log-likelihood for the fitted model with p covariates . We quantified LD between all pairs of biallelic SNPs using the absolute unsigned value of Lewontin's D′ statistic [73] . P values represent significance of departure from the null hypothesis that the pair is in equilibrium . All P values are two-tailed . Haploview was used for the LD plots . Three-dimensional models of the two CCRL2 proteins with 167F or 167Y were constructed using the PROTINFO structure prediction server ( http://www . protinfo . compbio . washington . edu ) , using the comparative modeling protocol [74] , [75] . The detailed modeling method is presented in Text S1 . The CCRL2 TM topology was established using ConPred II ( http://bioinfo . si . hirosaki-u . ac . jp/~ConPred2/ ) , a predication program based on consensus results of several prediction methods including TMpred , TMAP , TMHMM , HMTOP and MEMSAT [50] . cDNAs coding for the full-length open reading frame ( ORF ) of human CCRL2 carrying 167Y or 167F were PCR amplified from two individuals with the respective homozygous genotype with proofreading DNA polymerase pfu ( Stratage , La Jolla , CA ) . After confirmation of sequence accuracy , they were ligated into the pcDNA3 ( Invitrogen , Gaithersburg , MD ) . Human embryonic epithelial cells line HEK293 was transfected with the constructs of CCRL2-167F or CCRL2-167Y . Stable transfectants were selected by culturing the cells in 800 µg/ml G418 . Once the stable cell lines were established , they were examined for chemotactic responses to chemokines . Migration of CCRL2-167F and CCRL2-167Y transfected HEK293 cells was assessed using a 48-well microchemotaxis chamber technique . They were examined for chemotactic responses to the following chemokines: CCL5 , CCL3 , CCL4 , CCL2 , CCL8 , CCL7 , CCL11 , CXCL12 , CCL21 , CCL20 , and CXCL10 . The cells were also tested for migration in response to chemotactic peptides using formayl peptide receptors , including W peptide and MMK-1 ( Text S1 ) . A cDNA coding for 169Y-CCR3 was created from the wild-type human 169F-CCR3 ligated into the pcDNA3 , which were used to transfect HEK293 cells ( Text S1 ) . | Human chemokine receptors are cell surface proteins that may be utilized by HIV-1 for entry into host cells . DNA variation in the HIV-1 major coreceptor CCR5 affects HIV-1 infection and progression . This study comprehensively assesses the role of genetic variation of multiple chemokine receptor genes clustered in the chromosome 3p21 and 3p22 on HIV-1 disease outcomes in HIV-1 natural history cohorts . The multivariate survival analyses identified functional variants that altered disease progression rate in CCRL2 , CCR3 , and CCR8 . CCRL2-F167Y affects the rate to AIDS development through a specific protection against pneumocystis pneumonia ( PCP ) , a common AIDS–defining condition . Our study identified this atypical chemokine receptor CCRL2 as a key factor involved in PCP , possibly through inducing inflammation in the lung . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"genetic",
"association",
"studies",
"hiv",
"genetics",
"biology",
"human",
"genetics",
"viral",
"diseases",
"genetics",
"and",
"genomics"
] | 2011 | Role of Exonic Variation in Chemokine Receptor Genes on AIDS: CCRL2 F167Y Association with Pneumocystis Pneumonia |
Numerous problems encountered in computational biology can be formulated as optimization problems . In this context , optimization of drug release characteristics or dosing schedules for anticancer agents has become a prominent area not only for the development of new drugs , but also for established drugs . However , in complex systems , optimization of drug exposure is not a trivial task and cannot be efficiently addressed through trial-error simulation exercises . Finding a solution to those problems is a challenging task which requires more advanced strategies like optimal control theory . In this work , we perform an optimal control analysis on a previously developed computational model for the testosterone effects of triptorelin in prostate cancer patients with the goal of finding optimal drug-release characteristics . We demonstrate how numerical control optimization of non-linear models can be used to find better therapeutic approaches in order to improve the final outcome of the patients .
Optimizing delivery systems targeting constant levels of drug concentration represents always a challenge for chronic diseases requiring continuous treatment and especially in those cases where the relationship between drug exposure ( represented generally as levels of drug concentration plasma measured longitudinally ) and pharmacological response is complex and non-linear . The management of prostate cancer with sustained release formulations of triptorelin ( TRP ) injected every 3-6 months represents a good example [1] . For the case of the hormone-sensitive prostate tumors the therapeutic goal of any pharmacology treatment is to maintain as longer as possible the levels of testosterone ( TST ) below the castration limit ( CT ) which is set to the plasma concentration value of 0 . 5 ng/mL [2] . In recent past , we have developed a mechanistic computational model for the TST effects of the agonist TRP in prostate cancer patients using longitudinal pharmacokinetic ( PK; drug concentration in plasma ) and pharmacodynamics ( PD; TST concentrations in plasma ) data obtained from several clinical trials testing the efficacy of different sustained-release formulations ( SR ) [3] . Briefly , TRP exerts its action by increasing the fraction of activated receptors and therefore stimulating the production of TST . However , the prolonged exposure of TRP causes receptor down-regulation , resulting in a reduced synthesis of TST . The typical TST vs time profile after a single injection of TRP is represented in Fig 1 . The schematic representation of the PKPD model developed for TST effects of TRP , excluding the absorption compartments of the original model , and the estimates of model parameters are shown in Fig 2 . As highlighted in Fig 1 there are three critical aspects to be taken into consideration at the time to develop an innovative delivery system of TRP for the treatment of prostate cancer: initial flare up , time to reach CT , and castration period . Ideally , such new formulation should release TRP at a rate eliciting levels of concentration in plasma minimizing both the initial flare up and the time to reach CT , as well as maximizing the castration period . Specifically , limitation in the TST flare-up ( TSTmax ) to 50% increase with respect to baseline , minimize time to castration after first injection ( tcast ) to values below 3 weeks , and extend the castration time after injection ( teffect ) for at least 9 months . Given the complex relationship between concentrations of TRP in plasma and response as represented in Figs 1 and 2 , together with the requisite of maintaining the TST profiles within the constraints mentioned above , optimization of the rate of drug release is not a trivial task and cannot be efficiently addressed through an extensive trial & error simulation exercise . In the current work we aimed to optimize the release profile of TRP from SR formulations matching the multi-objective therapeutic needs applying optimal control methodology [4] . The rationale behind the decision of focusing on the release process is based on the assumption that once the drug is absorbed and reaches systemic circulation ( represented as part of the central compartment in Fig 2 ) it follows the same distribution and elimination characteristics regardless the type of formulation administered . The same is assumed with respect to the TST response , the rate of synthesis and degradation of TST and receptors , the dynamics of receptor occupation , and the down-regulation process . These mechanisms are independent from the absorption properties of the drug . Despite we focused on a specific case , the workflow and methodology used can be readily translate to other therapeutic areas and scenarios such as dosing schedule optimization and personalized treatments .
Values listed in the table inserted in Fig 2 include estimates of typical population parameters and between-subject variability ( represented by BSV in the table and hereafter ) obtained from [3] for a set of model parameters . In order to obtain the population of virtual patients , parameters were modelled as P i = P p o p × e η i _ P , where Pi and Ppop represent the ith individual and typical population values of the P parameter , respectively , and ηi_P corresponds the deviation of Pi with respect the typical value Ppop; the set of individual ηi_P forms a random variable with mean value of 0 and variance ω P 2 following a normal distribution , whereas the distribution of individual parameters is log-normal . The magnitude of ω P 2 reflects the BSV associated to a specific model parameter , which in Fig 2 is expressed as coefficient of variation ( CV% ) . One thousand set of disposition ( clearance and volume of distribution in the central compartment , represented as CL and Vc respectively ) , pharmacodynamics ( receptor equilibrium dissociation constant of triptorelin ( KD ) ) and system ( baseline TST levels ( TST0 ) , zero-order rate of TST production independent from gonadotropins ( kin ) , zero-order rate constants of receptor synthesis ( kS_R ) and the value that elicits a 50% maximal reduction in kS_R for a given amount of total receptors ( DR_50 ) ) related parameters were generated using the typical population estimates and corresponding marginal distributions reported in the table of Fig 2 . The parameter values for the virtual population were generated with NONMEM 7 . 2 [5] . An optimal control problem is a dynamic optimization problem in which the state of a system is linked in time to the application of a control function u ( t ) , which drives the system towards a desirable outcome by minimizing a cost function J ( u ) subject to operating constraints [4 , 6] . In other words , the control variable u ( t ) forces a system to have an optimal performance . The concrete control strategy will depend upon the criterion used to decide what is meant by “optimal”; in the current case TSTmax < 1 . 5 ⋅ TST0 ng/ml , minimize tcast ≤ 3 weeks , and maximize teffect ≥ 9 months . Therefore any optimal control problem can be formulated to find the magnitude of u ( t ) over the time of study [from initial time t0 to final time tf] such that: min u ( t ) J ( u ) = ϕ [ x ( t f ) ] + ∫ t 0 t f L [ x ( t ) , u ( t ) ] d t subject to d x ( t ) d t = f ( x ( t ) , u ( t ) , t ) x ( t 0 ) = x 0 , h ( x ( t ) , u ( t ) ) = 0 , g ( x ( t ) , u ( t ) ) ≤ 0 , ( 1 ) where J ( u ) is the cost function , u ( t ) is the control variable; x ( t ) the vector of state variables; x0 the set of initial conditions of the state variables; h ( ) the equality constraints; and g ( ) the inequality constraints . The general form of the equation in J ( u ) is known as Bolza optimization problem [7] , which is represented as the sum of a terminal cost functional ( Mayer problem ) and an integral function of the state and control from t0 to tf ( Lagrange problem ) . For a more detailed information see S1 Text . Fig 2 shows a schematic representation of the state variables and control input defined in this work . The state system is characterized by the variables that predict serum concentrations of triptorelin ( CTRP ) , concentrations of triptorelin in the shallow and deep peripheral compartments ( C1 and C2 respectively ) , drug input profile ( D ) , amount of total receptors ( RT ) , and optimal testosterone levels ( TST ) , each of them represented by the corresponding ordinary differential equation as shown in Table 1 . Note that in Fig 2 the terms resembling the 0th and 1st order absorption processes have been removed from the original model structure from [3] and have been replaced by the new control variable u ( t ) . Therefore the expression associated to the rate of change of the levels of TRP in plasma ( C TRP ( t ) ˙ ) is: C T R P ˙ = u ( t ) + C L D 1 V T 1 · c 1 ( t ) + C L D 2 V T 2 · c 2 ( t ) - C L D 1 V c · C T R P ( t ) - C L D 2 V c · C T R P ( t ) - C L V c · C T R P ( t ) ( 2 ) An additional compartment D was defined , where the dose of TRP administered to the patients ( 10mg in this evaluation exercise ) was placed as initial condition ( D0 ) . The control variable u ( t ) leaves this compartment and enters to the central ( systemic ) compartment of TRP as follows: D ˙ = - u ( t ) ( 3 ) Recall that , u ( t ) ( ng/day ) does not represent any particular mechanism of absorption ( i . e . , zero and/or first order kinetics ) , but a vector of different values that influence the system to behave in a pre-determined ( optimal ) way . The aim of this work was to find the time profile of u ( t ) ( input function of TRP into the central compartment ) that minimizes an objective ( or cost ) function and satisfies all constraints which represent the boundaries and therapeutic goals to be achieved ( see Table 1 ) . The choice of an objective function represents a critical aspect in optimal control problems [8] . Here , the problem is divided into two phases each represented by a different cost function and defined between: ( i ) 0 and tcast , and ( ii ) tcast and ≥ 280 + tcast days , respectively . During the first phase ( from 0 to tcast ) the u ( t ) profile is optimized to transfer the system from an initial state TST0 ( baseline testosterone level ) to the final state of 0 . 5ng/ml ( CT value ) in the shortest possible time . To solve the first phase of the optimization problem , the following objective function and equality constraint were defined respectively: J I ( u ) = t c a s t ( 4 ) T S T ( t = t c a s t ) - 0 . 5 = 0 ( 5 ) where tcast is an static control variable for minimizing JI . Here , we wished to obtain the minimum value of tcast that causes TST levels to achieve the CT value . The final time tcast was not known in advance , and that is the reason why the optimization problem was divided into two different phases . Additionally , an inequality constraint was added to limit the initial flare-up of the testosterone below 50% increase with respect to baseline: T S T ( t ) < 1 . 5 · T S T 0 ( 6 ) The second phase , covering the period between tcast and 280+tcast days , aims to maintain the TST levels below CT . If a second objective function or constraints were not incorporated into the optimization problem , values of TST rose above CT at times much earlier than 280 days . The approach used to overcome the above mentioned undesired effect and maintain TST predictions within the therapeutic goal led to the minimization of a second objective function of the form: J I I ( u ) = ∫ t c a s t 280 + t c a s t T S T ( t ) 2 d t ( 7 ) The rationale for formulating JII using a quadratic function ( TST ( t ) 2 ) for minimizing testosterone levels , instead of TST ( t ) , was because it offers relevant mathematical advantages in the context of optimization problems . In optimal control theory , one of the main necessary conditions for optimality is that control variables minimize a Hamiltonian function over u ( t ) . The Hamiltonian becomes convex if quadratic forms are used for the objectives and thus the problem will have a unique minimizer [4] . See S1 Text for more information about the Hamiltonian matrix and the necessary and sufficient conditions for optimal control problems . Furthermore , using squared terms amplify the effects of large variations and de-emphasize the contributions of small fluctuations . Continuity between the two phases of the optimization problem was ensured by imposing the initial conditions of the state variables at phase II to be equal to their final values at the end of the phase I ( see Table 1 ) . Alternatively , other objective functions or constraints could have been defined . For example , an alternative approach to model the second phase of the optimal control problem is to only add the inequality constraint TST[tcast: ( 280 + tcast ) ] − 0 . 5 < 0 , instead of a second objective function JII . This approach resulted in TST levels closer to CT compared to the values obtained with the addition of JII . However , in the work from [1 , 9] suggested that a CT value lower than 0 . 2 ng/ml could be an even better target to maximize therapeutic outcomes of prostate cancer patients . Due to these variations in the definition of the most appropriate CT value , we prioritized the minimization of TST levels during the second phase using JII because we obtained the lowest possible values of TST . Table 1 summarizes the setup of the different components of the optimal control problem described above . There exists different methods to solve this type of problems [10 , 11] . In our case , the dynamic optimization problem was solved numerically via direct methods with the IPOPT Solver ( Interior Point OPTimizer ) [12] which is freely available in the APMonitor Optimization Suite ( http://apmonitor . com/ ) through MATLAB programming environment [13] . The results were evaluated by calculating the proportion of individuals that achieved the described therapeutic goals and constraints . For more information about control theory see the works from [6 , 8 , 14] and for a more comprehensive overview of the role of optimal control in cancer research read the reviews from [4 , 10 , 15] . During the optimal control exercise , values of TST in plasma were obtained approximately every 12h for the first phase and every 120h in the second phase . Given the fact that the disposition , pharmacodynamics , and system parameters were already known as they were randomly generated as described in Material and methods , the analyses of the TST profiles described in this section focused on the mechanistic/parametric characterization of the absorption process of TRP aiming to provide biopharmaceutics with metrics useful to guide the development of new sustained release formulations . Those metrics are the fractions of the total dose injected absorbed following 0th and 1st order processes , the cumulative drug release profiles over time , the percentage of the dose that should remain in the site of injection at tcast and the time at which the different absorption mechanism are activated . The absorption model used to estimate the corresponding absorption parameters allowing afterwards computation of metrics is represented in the work from [3] and comprises three non-simultaneous absorption mechanisms , two of them following 1st order kinetics and the third one following a 0th order process . This model is considered of a sufficient complexity to deal with almost any absorption profile that can take place after administration of SR formulations [16 , 17] . A schematic representation of the structural model with the corresponding ordinary differential equations is provided in S1 Fig . The analyses were performed with the NONMEM version 7 . 2 software [5] , following a two stage approach in which the parameters of each subject are first obtained and summary statistics ( median , and 95th confidence intervals ) are then calculated . BSV in the absorption parameters was modelled exponentially as described in section for the rest of model parameters . TST concentrations obtained in step were logarithmically transformed for the analysis , and residual variability was modeled by using an additive error model on log-transformed data .
Fig 4 ( blue points ) illustrates the optimal testosterone profiles for the 1000 hypothetical individuals that we obtained after applying the optimal control problem formulated in Table 1 . The initial dose was considered to be 10mg . The code and data to reproduce these results in MATLAB can be found in the S1 Data . All of them achieved the 3 quantitative therapeutic goals ( 95% interval confidence between parenthesis ) defined in the Introduction section: time to castration was minimized to 18 . 96 days ( 11 . 408—36 . 289 ) , the increase of TST levels at the flare was always smaller than 50% with respect to baseline ( 36 . 8%-50 . 002% ) , and teffect was greater than 280 for all the patients . These profiles were generated with the manipulable variable u ( t ) which could take any values in order to minimize the multi-objective problem . However if we looked to the TRP concentration vs time profiles that induced the optimal TST levels ( data not shown ) , those profiles did not seem attainable by using simple first or zero order kinetics . That was the reason to directly approximate the TST levels with the PKPD model presented by [3] and estimate the most adequate absorption parameters . The optimal release characteristics corresponding to the selected PKPD model from [3] are listed in Table 2 . The final model adequately described the optimal TST profiles calculated in the previous section as shown in the individual profiles of Fig 5 . A lag time was associated with one absorption compartment . The first order rate constant of absorption of the second depot compartment ( KA2 ) had a very low median value ( 0 . 003 day−1 ) , resulting in a slow decay of TRP in serum concentrations . The first order rate constant of the first depot compartment ( KA1 ) , instead , had a higher value ( 0 . 25 day−1 ) to allow for a rapid decay of the TST levels in the firsts days of treatment . Table 2 also indicates that most of the drug is released following 1st order kinetics as the fraction of drug associated with the 0th order absorption process ( Finf ) is very small ( 4% ) . This result is also reflected in Fig 6 , where the median tendency of the drug release following each of the absorption mechanisms for the 1000 individuals is shown . The values of the duration of the 0th order process ( Dinf ) varied immensely between individuals ( from hours to more than 200 days ) , thus the variability term was removed from this parameter . The therapeutic objectives obtained were compared to those from the optimal TST profiles ( Fig 7 ) and the 95% confidence intervals were also calculated from the 1000 samples . Minimal time to achieve castration levels was 19 . 5 days ( 11 . 4-56 . 7 ) and the median percentage of drug consumed until that moment was 38 . 9% ( 16 . 55%-66 . 96% ) . In Fig 7A the distribution of tcast values for the 1000 individuals can be appreciated . With the modeling approach 63 . 9% of the patients had a tcast smaller than 21 days , whereas in the optimal TST profiles this value was equal to 70 . 7% . Regarding the second therapeutic goal , the initial peak in the TST levels had a median of 55% ( 17%-75% ) increase with respect to baseline . This indicated that the second objective was not always achieved , contrary to the case of the TST profiles obtained by the optimal control problem where the flare up was much more controlled ( Fig 7B ) . Nonetheless , the median value was very close to the optimal value of 50% , so we assumed that the modeling approach managed to achieve the second therapeutic goal as well . Finally , the long-term castration had a median value of 351 days ( 235 . 9—708 ) , which was higher than expected ( Fig 7C ) , but we again needed to take into account that a small fraction of individuals ( 7 . 8% ) did not achieve a tcast + teffect above 9 months due to their specific physiological characteristics .
Optimal control theory has been applied to a population pharmacokinetic/pharmacodynamic model to derive the optimal drug release profiles to achieve multiple therapeutic goals . The optimal control analysis is more relevant in physiological systems with complex dynamics where simple simulation tuning parameters exercises are not effective to obtain the optimal profiles . Moreover , the flexibility of the method allows to deal with multiple and tight therapeutic objectives performing real optimization . In this context the question of how to define the objective functions and how to quantify our therapeutic goals becomes crucial . Here , we focused on the resulting testosterone levels of the patients , however , within the oncology area , different therapeutic objectives can be established with the goal of improving drug combinations , help to lessen the side effects of cancer treatments , etc . Finally , the optimal release characteristics have been described based on standard absorption PK models . Although there are some discrepancies between the resulting TST profiles from the optimal control strategy and the modeling approach ( see Results ) , we note that the important aspect of this work was to find the optimal release characteristics for prostate cancer patients , not to perform an ideal PKPD modeling exercise as there was not real data to fit . We conclude that this objective is achieved and that the information summarized in this article could be very useful for the development of new formulations , since it provides insight into the desired absorption characteristics and could produce a broad benefit for future prostate cancer patients . | Mathematical models of the disease processes are widely used in computational biology to quantitatively describe the time course of disease progression and are often linked to pharmacokinetic–pharmacodynamic models in order to evaluate the effect of drug treatment on disease . Once the models are built from observed information and/or literature data , they can predict the dynamics of the system under different conditions through computer simulations . However , simulation exercises are not always effective to obtain the desired objectives due to the complexity of these systems . In this work , we optimized the release characteristics of a synthetic gonadotropin-releasing hormone analog used to induce chemical castration by inhibiting the testosterone levels in prostate cancer patients . The therapeutic goals to achieve were to minimize the initial flare up of testosterone levels and the time to reach testosterone values below castration limit , while maximizing the castration period of the patients . Our methodology , based on control theory , introduces a manipulable variable into the system’s equations to drive the model towards the established goals . We demonstrated how drug-release properties can be improved with the implementation of optimal control strategies to enhance the outcome of cancer patients . These methods are extrapolable to other problems encountered in the field . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"urology",
"medicine",
"and",
"health",
"sciences",
"drug",
"absorption",
"cancer",
"treatment",
"sorption",
"cancers",
"and",
"neoplasms",
"genitourinary",
"tract",
"tumors",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"hormones",
"castration",
"oncology",
... | 2018 | Optimal dynamic control approach in a multi-objective therapeutic scenario: Application to drug delivery in the treatment of prostate cancer |
Enveloped viruses need to fuse with a host cell membrane in order to deliver their genome into the host cell . While some viruses fuse with the plasma membrane , many viruses are endocytosed prior to fusion . Specific cues in the endosomal microenvironment induce conformational changes in the viral fusion proteins leading to viral and host membrane fusion . In the present study we investigated the entry of coronaviruses ( CoVs ) . Using siRNA gene silencing , we found that proteins known to be important for late endosomal maturation and endosome-lysosome fusion profoundly promote infection of cells with mouse hepatitis coronavirus ( MHV ) . Using recombinant MHVs expressing reporter genes as well as a novel , replication-independent fusion assay we confirmed the importance of clathrin-mediated endocytosis and demonstrated that trafficking of MHV to lysosomes is required for fusion and productive entry to occur . Nevertheless , MHV was shown to be less sensitive to perturbation of endosomal pH than vesicular stomatitis virus and influenza A virus , which fuse in early and late endosomes , respectively . Our results indicate that entry of MHV depends on proteolytic processing of its fusion protein S by lysosomal proteases . Fusion of MHV was severely inhibited by a pan-lysosomal protease inhibitor , while trafficking of MHV to lysosomes and processing by lysosomal proteases was no longer required when a furin cleavage site was introduced in the S protein immediately upstream of the fusion peptide . Also entry of feline CoV was shown to depend on trafficking to lysosomes and processing by lysosomal proteases . In contrast , MERS-CoV , which contains a minimal furin cleavage site just upstream of the fusion peptide , was negatively affected by inhibition of furin , but not of lysosomal proteases . We conclude that a proteolytic cleavage site in the CoV S protein directly upstream of the fusion peptide is an essential determinant of the intracellular site of fusion .
To achieve successful infection enveloped viruses need to fuse with a host cell membrane to deliver the viral genome into the host cell . Some viruses , such as herpes simplex virus , Sendai virus , and human immunodeficiency virus , appear to be capable of direct fusion at the plasma membrane after initial attachment [1]–[5] . However , the majority of enveloped viruses use endocytosis for uptake and transport prior to fusion . Since endocytic cargo may eventually end up in the destructive environment of the lysosome , environmental cues are crucial to trigger viral fusion at the right stage of trafficking . These triggers , which may include a decrease in pH , changes in redox environment , and proteolytic activity [6]–[8] , induce conformational changes in the viral fusion proteins leading to the merger of viral and host membranes . Two well-studied viruses; influenza A virus ( IAV ) and vesicular stomatitis virus ( VSV ) , are known to undergo fusion upon exposure to low pH [9]–[12] . Other enveloped viruses , such as respiratory syncytial virus ( RSV ) and Ebola virus , require proteolytic processing of their viral fusion proteins in the endosomal system for fusion to occur [13]–[16] . Coronaviruses ( CoVs ) are enveloped , plus-strand RNA viruses belonging to the family Coronaviridae in the order Nidovirales . They are capable of infecting a wide variety of mammalian and avian species . In most cases they cause respiratory and/or intestinal tract disease . Human coronaviruses ( HCoVs ) are known as major causes of the common cold ( e . g . HCoV-229E and HCoV-OC43 ) . However , the emergence of new HCoVs of zoonotic origin has shown the potential of CoVs to cause life-threatening disease in humans as was demonstrated during the 2002/2003 SARS-CoV epidemics and more recently for MERS-CoV in the Middle East [17] , [18] . The well-studied mouse hepatitis virus ( MHV ) is often used as a safe model to study CoV infections . All CoV virions contain a canonical set of four structural proteins . The viral genomic RNA is encapsidated by the nucleocapsid protein ( N ) to form the helical nucleocapsid , which is surrounded by the lipoprotein envelope , containing membrane glycoprotein ( M ) , the small envelope protein ( E ) , as well as the spike glycoprotein ( S ) ( reviewed in [19] ) . Trimers of the CoV S protein , a type I membrane protein belonging to the class I fusion proteins , form the peplomers that protrude from the virion surface [20] . The S protein can be divided into two functional subunits . The amino-terminal S1 subunit contains the receptor-binding domain; while the carboxy-terminal S2 subunit contains domains required for fusion , including the fusion peptide ( FP ) , heptad repeat domains ( HR ) HR1 and HR2 , and the transmembrane ( TM ) domain . Various entry routes have been described as being used by different CoVs for infection of cells . Clathrin-dependent as well as clathrin- and caveolae-independent entry pathways have been reported for SARS-CoV [21] , [22] . Also feline infectious peritonitis virus ( FIPV ) was suggested to enter via a clathrin- and caveolae-independent endocytic route [23] , [24] . For the HCoV-229E a caveolae-dependent endocytic uptake has been suggested [25] . Although the ability of MHV S proteins to cause cell-cell fusion at a neutral pH was initially interpreted as an indication for fusion of virions at the cell surface , more recent studies indicate the requirement for clathrin-mediated endocytosis for entry of MHV [26]–[29] . However , while some studies report that MHV strain A59 is sensitive to lysosomotropic agents that affect endocytosis [26] , this is not the case according to others [27] . Proteolytic cleavage of the CoV S proteins appears to be important for the induction of cell-cell fusion and/or virus entry into host cells . Different cleavage sites have been identified for different CoVs , the importance of which seems to differ for cell-cell and virus-cell fusion . Some CoV S proteins , including that of MHV strain A59 , are cleaved at the S1/S2 boundary by furin ( -like ) proteases during transport of the newly assembled virions through the secretory pathway of the producer cell [30]–[33] . Inhibition of this S protein cleavage was shown to inhibit cell-cell fusion , but not to affect entry of MHV strain A59 into host cells [30] , [34] , [35] . MHV strain 2 contains an S protein that is not cleaved at the S1/S2 boundary . Interestingly , although MHV strains 2 and A59 were both reported to enter via clathrin-mediated endocytosis , entry of MHV 2 but not of MHV A59 , was blocked by inhibitors of low-pH activated cathepsin proteases [27] , [36] . Inhibitors of cathepsin proteases have also been shown to inhibit entry of SARS-CoV and feline CoVs [23] , [37] , [38] , while treatment of cell-bound virus particles with different proteases was shown to enhance virus entry and/or cell-cell fusion [27] , [34] , [39]–[45] . For SARS-CoV and infectious bronchitis virus ( IBV ) , it appears that a proteolytic cleavage of the S protein at a more downstream position than the S1/S2 boundary upon receptor binding is of importance for cell entry [40] , [43] , [46]–[49] . In the present study we performed a detailed investigation of the entry of different CoVs . Using siRNA gene silencing , we found that the prototypic coronavirus MHV strain A59 ( further referred to as MHV ) requires proteins known to be important for late endosomal maturation and endosome-lysosome fusion for efficient infection of cells . By using recombinant MHVs expressing reporter genes as well as by applying a novel , replication-independent fusion assay we confirmed the importance of clathrin-mediated endocytosis and demonstrated that trafficking of MHV virions to lysosomal compartments and processing of the S protein by lysosomal proteases was required for productive entry to occur . Our results indicate that a cleavage site in the S protein of CoVs immediately upstream of the FP determines the site of fusion . In agreement herewith FIPV , which requires processing by lysosomal proteases , was also shown to depend on trafficking to lysosomes . In contrast , MERS-CoV , which contains a minimal furin-cleavage site consensus sequence in the S protein immediately upstream of the FP , was negatively affected by inhibition of furin , but not of lysosomal proteases .
In an automated , high-throughput RNAi screen [50] targeting the druggable genome ( approximately 7000 genes ) a number of proteins associated with endocytosis were found to be required for efficient infection of HeLa cells with GFP-expressing MHV . To validate these findings these proteins were subjected to a follow-up analysis using siRNA-mediated gene silencing with oligonucleotides from a different supplier than the one used for the initial RNAi screen ( Fig . 1A ) . The follow-up analysis included ACTR2 and ACTR3 , two major constituents of the Arp2/3 complex which are important for the formation of actin branches and cell surface protrusions , as well as for the motility of several pathogens inside host cells ( reviewed in [51] , [52] ) . Also selected were the RAS-related GTP-binding protein family members , RAB7A and RAB7B , which have been shown to be involved in endosomal maturation ( reviewed in [53] ) . RAB7 interacts amongst others with members of the homotypic fusion and vacuole protein sorting ( HOPS ) tethering complex , involved in late endosome to lysosome maturation . The HOPS subunit VPS39 ( reviewed in [54] ) was also found to be a strong hit in the siRNA screen and therefore selected . Other proteins included SNX1 , involved in retrograde transport of cargo between endosomes and the trans-Golgi network ( reviewed in [55] ) , VCL , inter alia involved in connecting the Arp2/3 complex with integrins during actin polymerization ( reviewed in [56] ) , and the Ser/Thr-protein kinase PAK1 , which is activated by the Rho/Rac/Cdc42 family and is implicated in a variety of downstream effects including modulation of the actin cytoskeleton ( reviewed in [57] ) . Transfection of HeLa cells carrying the receptor for MHV ( HeLa-mCC1a cells ) with different siRNAs was followed by an infection with GFP-expressing MHV ( MHV-EGFPM ) at low multiplicity of infection ( MOI ) , resulting in approximately 10–15% infected cells under control conditions . After 8 h of infection cells were collected and GFP expression by the replication of MHV was analyzed by fluorescence-activated cell sorting ( FACS ) . As controls siRNAs silencing GFP and negative-control siRNAs were used . A hit from the screen was considered as confirmed when transfection with at least two out three independent siRNAs resulted in significant reduction in MHV-driven GFP expression relative to the negative-control siRNAs . siRNA-mediated gene silencing of ACTR2 and ACTR3 resulted in reduced infections for all three siRNAs , indicating that actin branching is important for MHV infection ( Figure 1A , dark orange ) . Also the importance of the RAB7A , RAB7B and VPS39 proteins , involved in late-endosome and late-endosome to lysosome maturation , for MHV infection could be confirmed ( Figure 1A , turquoise and light green ) . The importance of SNX1 , VCL and PAK1 for infection of HeLa cells with MHV could not be confirmed ( Figure 1A , grey ) . The latter three genes were not studied any further . To validate our transfection protocol and confirm the efficacies of the siRNAs at the mRNA level , quantitative RT-PCR analysis was performed . All siRNAs used reduced the corresponding mRNA levels with 75–95% ( Figure 1B ) . siRNAs targeting RAB7A were shown to inhibit the expression of a RAB7a-fusion protein ( Figure S1 in Text S1 ) . To confirm and extend our understanding of the role of endocytosis in MHV entry we subsequently selected a number of proteins known to be involved in either caveolae- or clathrin-mediated endocytosis , actin- or microtubule-mediated transport , as well as proteins associated with endosomal vesicles and endosomal maturation , to be screened using the siRNA silencing-approach described above . Again , proteins were considered important for infection with MHV when transfection with at least two out three independent siRNAs resulted in significant reduction in MHV-driven GFP expression relative to the negative-control siRNAs . siRNA-mediated downregulation of proteins involved in caveolae-mediated endocytosis revealed that CAV2 , but not the other proteins analyzed are important for infection with MHV ( Figure 1C , light blue ) . Downregulation of most proteins associated with clathrin-mediated endocytosis inhibited MHV infection , including DNM1 , DNM2 , CLTC , and DAB2 . siRNA-mediated silencing of EPS15 or AAK1 , accessory factors of clathrin-mediated endocytosis , did not affect MHV replication ( Figure 1C , dark blue ) . Silencing of early endosome-associated genes ( EEA1 , RAB5A , RAB5B , and RAB5C; Figure 1C , cerulean ) each decreased replication-mediated GFP expression . While downregulation of MYO6 , involved in actin-based motility , did not influence MHV infection ( Figure 1C , dark orange ) , our results indicate that the microtubule-associated motility proteins DYNC1H1 and DYNC2H1 are important for infection with MHV ( Figure 1C , orange ) . Silencing of NSF , required for transport from early to late endosomes [58] , or of the HOPS subunits VPS11 and VPS41 , which are involved in late endosome to lysosome maturation ( Reviewed in [54] ) , all resulted in severely reduced MHV infection ( Figure 1C , turquoise and light green , respectively ) . To further explore the endocytic route and factors involved in MHV infection we determined the effect of inhibitors on MHV infection . HeLa-mCC1a cells were treated with endocytosis-affecting agents for 30 min and then infected with luciferase-expressing MHV ( MHV-EFLM; [59] ) in presence of the inhibitors , after which the inhibitors were kept present until cell lysis . When cells were inoculated with MHV-EFLM in the absence of inhibitors , the inhibitors were added to the cells at 2 h post infection ( hpi ) to assess effects of inhibitors on post-entry steps . At 7 hpi cells were lysed and firefly luciferase expression levels were determined . Infection in the presence of the solvents dimethyl sulfoxide ( DMSO ) and methanol ( MeOH ) , as well as the known inhibitors of MHV RNA synthesis Brefeldin A ( BrefA , inhibitor of GBF1 ) [60] and MG132 ( proteasome inhibitor , probably also affects MHV entry; [61] ) were included as controls . MHV infection was not affected by addition of the solvents , whereas both MG132 and BrefA severely decreased luciferase expression regardless of the time of addition . Inhibition of endosome maturation with ammonium chloride ( NH4Cl ) , Bafilomycin A1 ( BafA1 ) , or Chloroquine ( Chloq ) severely diminished luciferase expression when the inhibitors were added prior to infection . Much smaller effects were observed when these drugs were added at 2 hpi , indicating that the inhibitors mainly affect MHV entry ( Figure 2 , deep sky blue ) . Similar effects were observed with known inhibitors of clathrin-mediated endocytosis; Chlorpromazine ( Chlopro ) , Monensin ( Mon ) , Dynasore , and Dyngo-4A ( Dyngo ) . All these compounds strongly decreased MHV replication-mediated luciferase expression when added early but not when added at 2 hpi ( Figure 2 , dark blue ) . The actin- and macropinocytosis-affecting drug EIPA , which inhibits the Na+/H+ exchanger NHE1 , led to reduced luciferase expression both when added prior to and after entry of MHV at 2 hpi . Actin cytoskeleton altering drugs Latrunculin A ( LatA ) , Jasplakinolide ( Jasp ) , Cytochalasin B ( CytoB ) , and Cytochalasin D ( CytoD ) , or the inducer of microtubule depolymerization Nocodazole ( Noc ) only decreased MHV infection when added early , indicating a role for the actin and microtubule cytoskeleton in entry but not RNA replication ( Figure 2 , dark orange and orange ) . Likewise U18666A , a cholesterol transport-affecting agent , which also prevents maturation of late endosomes [62] , had a strong inhibitory effect on MHV infection when added early ( Figure 2 , turquoise ) . Collectively , these results indicate an important role for clathrin-mediated uptake and for endosome- and endosome-to-lysosome maturation for MHV infection . The time-of-addition experiments with the different inhibitors indicated that particularly the entry step of the MHV infection cycle is negatively affected by perturbation of clathrin-mediated endocytosis or of endosome maturation . However , assays based on reporter gene expression driven by virus replication do not allow discrimination between virus entry and RNA replication when analyzing siRNAs or agents that also affect RNA synthesis . To unequivocally demonstrate the importance of clathrin-mediated endocytosis and endosome maturation for MHV entry , we therefore made use of a fusion assay we recently developed [63] . The assay is based on minimal complementation of defective β-galactosidase ( β-galactosidase ΔM15 ) with the short α-peptide [64] . MHV-αN , a recombinant MHV containing an N protein tagged with the α-peptide ( αN ) , is used to infect ΔM15−fragment expressing target cells . Upon fusion of the virion with a host cell membrane αN is released into the cytoplasm resulting in complementation of the defective β-galactosidase thereby reconstituting a functional enzyme . Conversion of the non-fluorescent substrate fluorescein-di-β-D-galactopyranoside ( FDG ) by β-galactosidase into green fluorophores fluorescein ( FIC ) can be measured by FACS or fluorescence microscopy ( Figure S2 in Text S1 ) . To analyze the effect of RNAi-mediated gene silencing on fusion , HeLa cells expressing the MHV receptor and the ΔM15−fragment ( HeLa-mCC1a-ΔM15 cells ) were transfected with individual siRNAs and inoculated with MHV-αN at 72 h post transfection . Before infection cells were pre-loaded with FDG by hypotonic shock . After 100 min incubation of cells with virus at 37°C , cells were collected and the amount of FIC generated as a results of enzyme complementation analyzed by FACS . The fusion assay showed that silencing of neither CAV1 nor CAV2 affected MHV fusion ( Figure 3A , light blue ) , even though reduction of CAV2 was shown to affect MHV infection ( Figure 1C ) . However , downregulation of clathrin-mediated endocytosis associated proteins DNM2 and CLTC lead to strongly decreased fusion , as did the lack of early endosome-associated factors RAB5B and RAB5C ( Figure 3A , dark blue and cerulean , respectively ) . Fusion was also affected by RNAi-mediated reduction of actin cytoskeleton-associated proteins ACTR2 and ACTR3 ( Figure 3A , dark orange ) , proteins known to be involved in late endosome ( RAB7A , RAB7B ) and late endosome-to-lysosome maturation ( VPS11 , VPS39 , and VPS41 ) ( Figure 3A , turquoise and light green ) . The importance of clathrin-mediated endocytosis and endosome maturation for MHV fusion was confirmed by analysis of endocytosis-affecting agents using the fusion assay . After pre-loading with FDG , cells were pre-treated with the inhibitors for 30 min at 37°C , after which cells were inoculated with MHV-αN in the presence of the agents , and analyzed by FACS as described above . As controls we included protein synthesis inhibitor cycloheximide ( CHX ) , MHV fusion inhibitor peptide HR2 ( HR2 , [20] ) , MG132 and BrefA . Fusion of MHV was not affected by the solvents or CHX , the latter confirming that this assay is independent of RNA replication and protein synthesis . MHV fusion was barely affected by replication inhibitor BrefA , whereas MG132 had a clear negative effect , in agreement with the conclusion drawn previously that MG132 inhibits entry of MHV as well as RNA synthesis [61] . Inhibition of endosomal maturation by NH4Cl , BafA1 and Chloq ( Figure 3B , deep sky blue ) or of clathrin-mediated endocytosis by Chlopro , Mon , and Dynasore ( Figure 3B , dark blue ) severely inhibited MHV fusion . Disturbance of the actin cytoskeleton by EIPA or by LatA , CytoB , or CytoD reduced fusion by 75–80% ( Figure 3B , dark orange ) , while interference with microtubule polymerization by Noc had a smaller effect ( Figure 3B , orange ) . Late endosomal maturation arrest caused by U18666A reduced fusion to approximately 10% ( Figure 3B , turquoise ) . In conclusion , the replication-independent fusion assay confirmed the importance of clathrin-mediated endocytosis and of endosome maturation for entry of MHV . The data indicate that late endosome-to-lysosome maturation is required for efficient entry and fusion . To confirm the importance of endocytic uptake and the association of MHV with endosomal compartments we performed live-cell confocal microscopy . To this end , sucrose density gradient-purified MHV virus was covalently labeled with the low-pH resistant dye DyLight 488 ( MHV-DL488 ) . HeLa-mCC1a cells were transfected with plasmids to express monomeric RFP ( mRFP ) fusion proteins of RAB5 , RAB7 , or LAMP1 . At 24 h post transfection , MHV-DL488 was bound to cells at 4°C for 90 min . Inoculation medium was replaced by warm medium containing trypan blue , which immediately shifts the emission spectrum of surface bound particles rendering them undetectable in the 505–530 nm channel unless they get endocytosed [65] . Cells were imaged using a spinning-disc confocal microscope acquiring z-stacks in 30 s intervals over 10 min time frames from 10–70 min post warming . Only low-level RFP fusion protein expressing cells were selected for analysis . Interestingly , MHV particles newly appeared even 60 min post warming , in agreement with the notion that MHV enters in an unsynchronized manner ( unpublished results ) . Co-localization and co-trafficking of viruses with endosomal compartments was assessed by detecting virus particles based on size and intensity ( green channel ) and by measuring the underlying intensity in the red channel ( endosomal vesicles ) . MHV virions were found to co-localize with all three endosomal compartments ( Fig . 4A ) . Whereas newly entering/appearing particles were always co-localizing with RAB5 molecules , they only associated with RAB7 and LAMP1 containing vesicles at later time points . To assess the association of MHV with endosomal vesicles during the entry process more extensively , we manually tracked the virus particles in the green channel and independently tracked the endosomal vesicles in the red channel in x/y and z-direction . A virion was categorized as associating with a certain endosomal marker only if this co-localization was observed over at least four sequential 30 s interval images . When the initial co-localization was lost , but the virion did not disappear , this virion was classified as associating/dissociating . Complete disappearance of a virus particle ( including in other z-stacks ) while immediately previously co-localizing with an endosomal marker was categorized as a fusion event ( Figures S3 and S4 in Text S1 ) . When a viral particle co-localized with endosomal compartments but did neither dissociate nor fade during the 10 min acquisition period it was classified as non-fusing . With this quantification method we analyzed 75–100 virions in total for each of the endosomal compartment types studied . The fraction of virions not fusing during the acquisition period was consistently found to be at around 10–15% . We observed that all of the entering MHV particles initially co-localized with RAB5-positive early endosomal vesicles and that most virions dissociated ( were no longer co-localized ) after 4–6 min . Notably , it appeared that in these events the RAB5 marker faded rather than moved away . Only a very small percentage of virions were categorized as fusing while in early endosomes . The number of fusion events was much higher for virions co-localizing with RAB7 or LAMP1 ( Figure 4B ) , indicating that most virions fuse in late endosomes or lysosomes . Our results so far indicate that most virions enter cells after having accessed late endosomes/lysosomes . We hypothesized that these compartments provide the environmental cues required for productive virus-cell fusion . In order to analyze to what extent the low pH in the endosomal system is required for entry of MHV , we analyzed the inhibition of MHV entry at different concentrations of BafA1 . While high concentrations of BafA1 ( as used for the results shown in Fig . 2 and 3 ) affect endosomal maturation , at low concentrations this inhibitor of vacuolar-type H+-ATPase only elevates the pH of endosomal compartments but does not affect endosomal trafficking per se [66] . We made use of that property and tested the sensitivity of MHV to BafA1 side by side with the control viruses VSV and IAV . VSV has been described to fuse at pH 6 . 2 in early and/or late endosomes [9] , [11] , [12] , [67]–[69] , while IAV has been shown to fuse in late endosomes at an even lower pH [9] , [10] , [70] . HeLa or HeLa-mCC1a cells were pretreated with increasing concentrations of BafA1 for 30 min prior to infection with reporter gene expressing viruses: VSV ( VSVΔG/FLuc-G*; [71] , [72] ) , IAV ( IAV-RLuc; [73] ) , or MHV ( MHV-EFLM ) . Luciferase expression levels indicated that infection of cells with VSV and IAV is much more affected by BafA1 , with an IC50 values of 0 . 80 and 0 . 63 nM , respectively , compared to MHV , which displays a three to four fold higher IC50 of 2 . 34 nM ( Figure 5A ) . Our results thus indicate that MHV is much less affected by perturbation of the endosomal pH than VSV and IAV . Nevertheless RNAi-mediated silencing of HOPS subunits and treatment of cells with U1866A indicates that late endosome-to-lysosome maturation is required for efficient entry . To confirm and extend these observations , we made use of haploid HAP1 cells lacking a functional HOPS complex resulting from lentiviral-mediated knockout of the VPS33A subunit ( H1-ΔV33 cells; [74] ) . Both HAP1 cells and H1-ΔV33 cells were modified to stably express the MHV receptor . As a control , the H1-ΔV33 cells were in addition stably transfected with FLAG-tagged VPS33A ( H1-ΔV33-fV33 ) . The different cells expressed similar levels of the MHV receptor as determined by FACS analysis ( Figure S5 in Text S1 ) . Expression of FLAG-VPS33A was confirmed by Western blot ( Figure S6 in Text S1 ) . Functional reconstitution was confirmed by confocal fluorescence imaging of lysosome localization ( Figure S7 in Text S1 ) . While in the knockout cells the lysosomes were clustered , the lysosomes were dispersed again throughout the cytoplasm in the FLAG-VPS33A re-transfected cells , as observed in the HAP1 parental cells . The haploid cells were infected with luciferase reporter gene-expressing MHV , VSV , or IAV at low MOI . Cells were lysed at 7 ( MHV and VSV ) or 16 ( IAV ) hpi and luciferase expression levels were determined . The lack of a functional HOPS complex had no effect on VSV and IAV infection; however , MHV infection was strongly reduced in the knockout , but not in the re-transfected cells ( Figure 5B ) . These observations confirm the conclusion that late endosome-to-lysosome maturation is required for efficient entry of MHV , a characteristic that is not shared with the pH-sensitive VSV and IAV . Considering that MHV was much less affected by perturbation of the endosomal pH than IAV and VSV while it requires trafficking to lysosomes for efficient entry , we hypothesized that entry might depend on cleavage of a viral protein by lysosomal proteases . Hence we analyzed the extent to which different protease inhibitors could inhibit MHV entry . Thus , HeLa-mCC1a-ΔM15 cells were pretreated for 30 min with the different inhibitors , after which the cells were inoculated with MHV-αN in inhibitor-containing medium . Cells were collected , loaded with FDG , and FDG conversion to FIC by complementation of β-galactosidase upon viral fusion was assessed by FACS . Our results indicate that most protease inhibitors tested ( Fig . 6 ) hardly inhibited fusion of MHV , if at all . Exceptions were AEBSF , which has been shown to cause aggregation of early endosomal vesicles [75] , and a pan-lysosomal protease inhibitor ( CPI; cystatin-pepstatin inhibitor ) capable of inhibiting the three major protease family members found in lysosomes . Thus , by using CPI we measured the combined effects of an endosomal papain-like cysteine protease inhibitor ( PLCP ) , an aspartyl protease inhibitor , and an asparagine endopeptidase inhibitor ( AEP ) [76] . From these results we conclude that inhibition of a broad range of endosomal proteases efficiently blocks fusion of MHV , indicating that efficient entry requires the activity of lysosomal proteases . In general , class I fusion proteins require cleavage just upstream of the FP to render them fusion competent [20] , [38] , [77] . However , while the S protein of MHV is cleaved at the S1/S2 boundary ( Fig . 7A ) , no protease cleavage site has been identified close to the fusion peptide . In view of the inhibition of MHV entry by the pan-lysosomal protease inhibitor CPI and in analogy to other class I fusion proteins , we hypothesized that an additional cleavage in the S protein , immediately upstream of the FP , is necessary to induce fusion . To test this hypothesis , we introduced an optimal furin cleavage site ( FCS ) by substituting three amino acids by Arg ( AIRGR→RRRRR ) immediately upstream of a highly conserved Arg ( indicated in bold ) that occurs just N-terminal of the FP . Recombinant MHV carrying this FCS in its S2 subunit was designated MHV-S2′FCS . ( Figure 7A ) . Western blot analysis of the S protein of a purified stock of this virus using an antibody recognizing the S2 subunit showed no evidence of cleavage at the newly introduced FCS ( S2′ site ) . Apparently , cleavage at this position does not occur during virus production ( Figure S8 in Text S1 ) . MHV carrying wild type or mutant S proteins displayed similar growth kinetics ( Figure S9 A and B in Text S1 ) . Next we analyzed whether the introduced FCS affected the sensitivity of the recombinant MHV to CPI , which does not exhibit inhibitory activity towards furin . Thus , HeLa-mCC1a cells were pretreated with CPI for 30 min and subsequently infected with wild type S ( MHV-EFLM ) or mutant S ( MHV-S2′FCS ) containing viruses expressing luciferase reporter genes in the presence of the protease inhibitor . At 7 hpi the cells were lysed and viral-replication dependent luciferase expression levels were determined . Introduction of the FCS resulted in the recombinant virus being no longer sensitive to inhibition by lysosomal proteases ( Figure 7B ) , probably because the S protein is now cleaved by furin in an endocytic compartment . To confirm that MHV-S2′FCS is no longer dependent on cleavage by lysosomal proteases , and to study its presumed dependence on furin cleavage for entry , we analyzed the ability of MHV-S2′FCS to infect the haploid cells that lack VPS33A - and thus the functional HOPS complex required for late endosome-to-lysosome maturation - in the absence or presence of furin inhibitor ( FI ) . After pretreatment of MHV receptor-expressing HAP1 , H1-ΔV33 , and H1-ΔV33-fV33A cells with furin inhibitor ( FI ) or mock treatment , cells were inoculated with MHV-EFLM or mutant virus MHV-S2′FCS in presence or absence of FI . At 7 hpi the cells were lysed and viral-replication dependent luciferase expression levels were determined . In agreement with previous results ( Fig . 5 ) , infection with MHV carrying a wild type S was severely reduced in cells lacking a functional HOPS complex and addition of the FI did not alter this effect ( Figure 8 , red bars ) . In contrast , infection with MHV-S2′FCS was not decreased by the lack of a functional HOPS complex . However , FI treatment had a clearly negative effect on this virus , which was much more dramatic in the absence of a functional HOPS complex in H1-ΔV33 cells ( Figure 8 , blue ) . In conclusion , MHV-S2′FCS lost the requirement for a functional HOPS complex in parallel with this virus becoming insensitive to the pan-lysosomal protease inhibitor CPI . In contrast to the virus with the wild type S , the mutant virus became sensitive to inhibition of furin cleavage . To explore MHV-S2′FCS entry requirements further we assessed the effect of RNAi-mediated downregulation of early and late endosome and HOPS complex associated genes . Therefore , HeLa-mCC1a-ΔM15 cells were transfected with each of three different siRNAs per gene for 72 h , after which they were infected with wild type ( MHV-EFLM ) or mutant ( MHV-S2′FCS ) S protein containing MHV . At 7 hpi the cells were lysed and viral-replication dependent luciferase expression levels were determined . As found previously ( Fig . 1 ) , infection with wild type S protein carrying MHV was reduced after gene silencing of RAB5 , RAB7 , VPS11 , and VPS41 ( Figure 9 , red bars ) . On the other hand , infection with MHV-S2′FCS was significantly diminished by downregulation of the early endosomal proteins RAB5B and RAB5C , but not of the late endosomal proteins RAB7A and RAB7B or the HOPS complex components VPS11 and VPS41 ( Figure 9 , blue bars ) . Consistently , infections with MHV carrying wild type or mutant S protein were equally blocked by inhibitors of clathrin-mediated endocytosis whereas the virus with the mutant S ( MHV-S2′FCS ) was much less sensitive to inhibitors of endosomal maturation , including BafA1 , or to perturbants of the actin cytoskeleton ( Figure S10 in Text S1 ) . From these results we conclude that introduction of a FCS immediately upstream of the FP abolishes the requirement for trafficking of virions to lysosomes and for processing by lysosomal proteases . The resulting virus , which still depends on clathrin-mediated endocytosis , now requires furin cleavage for efficient entry , the enzymes for which occur earlier in the endocytic pathway [78] . Our results indicate that the protease cleavage site upstream of the spike protein FP is an important determinant of the intracellular site of fusion . To gain more insight into the putative protease cleavage sites in the corresponding region of the S proteins of other CoVs , we analyzed the sequence of this region in several alpha , beta and gamma coronaviruses by performing ClustalW sequence alignment . The fusion peptide sequence was found to be highly conserved amongst the different coronaviruses . Also an Arginine residue immediately upstream of the predicted fusion peptide is highly conserved with the exception of FIPV ( serotype II ) . Interestingly , MERS-CoV and IBV-Beaudette contain a minimal furin cleavage site Arg-X-X-Arg just upstream of the fusion peptide ( Figure 10A ) . In analogy with the results obtained with FCS-mutant MHV , we predicted that FIPV and MERS-CoV would differ in their protease inhibitor sensitivity and lysosomal trafficking requirements . To corroborate these findings , we decided to analyze the entry of these two other coronaviruses . To this end , HeLa cells expressing the FIPV receptor ( HeLa-fAPN cells ) were subjected to siRNA-mediated downregulation of late endosomal proteins RAB7A and RAB7B or of HOPS complex subunits VPS11 , VPS41 , and VPS39 , followed by inoculation with luciferase expressing FIPV ( FIPV-Δ3abcRL; [79] ) . Infection with FIPV was significantly affected by siRNA-mediated downregulation of proteins required for late endosome-to-lysosome fusion ( Figure 10B ) . Since the requirement for a functional HOPS complex is indicative of fusion in lysosomes , as we observed for MHV , we analyzed whether FIPV requires processing by lysosomal proteases for efficient entry as well . The results indicate that this is indeed the case as FIPV-driven luciferase expression was diminished in the presence of the pan-lysosomal protease inhibitor CPI ( Fig . 10C ) . On the other hand , infection with FIPV was not affected by FI . As MERS-CoV carries a FCS in its S protein immediately upstream of the FP , we hypothesized this virus not to require trafficking to lysosomes and processing by lysosomal proteases for efficient entry . To test this prediction , Huh-7 cells were pretreated with FI or the pan-lysosomal protease inhibitor CPI for 30 min . Cells were subsequently inoculated with MERS-CoV at a MOI of 0 . 1 in the presence of these inhibitors . At 8 hpi the cells were fixed and the number of infected cells determined using immunocytochemistry and wide-field microscopy . The results indicate that , in contrast to wild type MHV and FIPV , but similarly to recombinant MHV carrying a FCS immediately upstream of the FP , infection with MERS-CoV is strongly inhibited by the FI but not by CPI ( Figure 11 ) , indicating that MERS-CoV does not require trafficking to lysosomes for efficient entry . Based on these results we conclude that the cleavage site in the CoV S protein immediately upstream of the FP is a key determinant of the intracellular site of fusion .
The results of this study provide an explanation for several , apparently conflicting results from earlier studies with respect to the process of MHV cell entry , particularly also regarding the necessity of proteolytic cleavage of the CoV S protein . By using a replication-independent fusion assay , we confirmed that MHV entry requires clathrin-mediated endocytosis despite the well-known ability of the MHV S protein to cause cell-cell fusion at neutral pH . We demonstrate that MHV particles traffic to and fuse in lysosomes . Yet , MHV is much less sensitive to perturbation of the low pH in the endo-/lysosomal system than low pH-dependent control viruses VSV and IAV . Our results additionally indicate that , for fusion to occur , the S protein of MHV requires proteolytic cleavage immediately upstream of the FP , like other class I fusion proteins . Efficient inhibition of MHV entry was only observed using a pan-lysosomal protease inhibitor , and could not be achieved using more specific protease inhibitors . Introduction of an optimal furin cleavage site in the S protein immediately upstream of the FP abolished the requirement for trafficking of virions to lysosomes for fusion . However , this virus still required clathrin-mediated uptake for efficient entry . Consistent with a common mechanism for the entry of CoVs , FIPV , but not MERS-CoV , the latter of which contains a furin cleavage site immediately upstream of the FP , was shown to require trafficking to lysosomes and processing by lysosomal proteases for efficient entry . Based on these results we propose a model in which the cleavage site immediately upstream of the FP is an essential determinant of the intracellular site of CoV fusion ( Figure 12 ) . The importance of clathrin-mediated endocytosis and endosomal trafficking in the entry of MHV was revealed by several complementary approaches . One of these was siRNA-mediated gene silencing . Although - with the exception of RAB7A - knockdown was not monitored at the protein level , we believe this approach firmly demonstrates the importance of novel host factors for several reasons . Validated siRNAs were used and the experimental conditions were confirmed by analyzing the mRNA expression levels of several genes by quantitative RT-PCR . Furthermore , we made use of three independent siRNAs per target gene , and a target was only classified as a hit when at least two out three siRNAs showed the same phenotype . Importantly , our findings were strengthened by targeting multiple proteins per host cell pathway/complex , each time with very similar results . Moreover , hits obtained with the replication-dependent reporter assays were confirmed with our novel replication-independent enzyme complementation entry assay . Also the use of recombinant viruses differing only in their spike proteins enabled us to show that inhibition of virus infection upon siRNA transfection resulted from differences in virus entry and not virus replication . Finally , the results obtained were corroborated by using a large panel of inhibitors and by making use of haploid knockout cells , in which late endosome-to-lysosome trafficking was inhibited . Our results demonstrate that MHV requires endocytic uptake for virus entry despite the S protein's ability to induce cell-cell fusion at neutral pH . Endocytic uptake is also required for a mutant virus carrying a S protein with a FCS immediately upstream of its FP , despite the relative insensitivity to high concentrations of BafA1 . Therefore , the ability of a virus to infect cells in the presence of BafA1 does not necessarily imply virus entry to occur at the cell surface . Also a recombinant MHV carrying the spike protein of MHV-4 ( MHV-JHM ) was found to enter via clathrin-mediated endocytosis ( MHV-S4; Figure S10 in Text S1 ) despite its ability to cause extensive cell-cell fusion [80]–[82] . The ability of MHV to cause cell-cell fusion at neutral pH while requiring endocytic uptake for virus-cell fusion suggests different requirements and triggers for these two fusion processes . Similarly , RSV was recently shown to enter cells after endocytic uptake despite the ability of this virus to cause cell-cell fusion [13] . The present study confirms and extends previous publications on MHV entry via clathrin-mediated endocytosis [26] , [83] . Both siRNAs downregulating clathrin-mediated endocytosis-associated proteins , such as clathrin heavy chain ( CLTC ) and Dynamin 2 ( DNM2 ) , and agents affecting this uptake pathway ( Chlopro , Dynasore , Dyngo-4a ) were capable of inhibiting infection with MHV . Importantly , these findings could be confirmed in our novel replication-independent virus-cell fusion assay , thereby directly showing an involvement of clathrin-mediated endocytosis in entry of MHV . Analysis of several accessory factors of clathrin-mediated endocytosis showed that clathrin-mediated entry of MHV strain A59 depends on clathrin-adaptor DAB2 , but not on EPS15 or AAK1 . Previously , clathrin-mediated entry of MHV strain 2 was also shown to be independent of EPS15 [83] . Based on the use of inhibitors , it was earlier concluded that MHV entry depends on cholesterol and lipid-rafts , which may be indicative of caveolae-mediated endocytosis [84] , [85] . Although our replication-dependent assays indicate a requirement for caveolin 2 ( CAV2 ) for infection , this protein was shown not to be involved in virus entry using our fusion assay . Also depletion of other proteins involved in caveolae-mediated endocytosis , including caveolin 1 ( CAV1 ) and flotillins 1 and 2 ( FLOT1 and FLOT2 ) did not affect MHV infection or fusion . Interestingly , fusion of MHV was severely inhibited by EIPA , an inhibitor of the Na+/H+ exchanger NHE1 , which is regarded as a hallmark inhibitor of macropinocytosis . Apparently , inhibition of virus entry by EIPA does not prove by itself that a virus enters via this particular pathway . EIPA has been reported to affect several other cellular processes , including actin remodeling , internalization of lipid rafts , distribution of endosomes , and even clathrin-mediated endocytosis [86]–[90] . Similar to the results obtained with the HeLa cells , also infection of murine LR7 cells was inhibited by compounds interfering with clathrin-mediated endocytosis ( Figure S11A in Text S1 ) . MHV virions require trafficking through the endocytic pathway to lysosomes for efficient entry . Upon clathrin-mediated uptake these virions are temporarily associated with early endosomes as demonstrated by co-localization with RAB5 during live cell imaging . Furthermore , the importance of early endosomes for entry was indicated by siRNA-mediated downregulation of various proteins associated with early endosomes ( EEA1 , RAB5A , RAB5B , and RAB5C ) , which inhibited MHV infection , as well as virus-cell fusion . However , only very few MHV particles appeared to fuse in the early endosomes . Live cell imaging indicated fusion largely to occur in late endosomes and/or lysosomes . Consistently , depletion of host proteins associated with late endosome and late endosome-to-lysosome maturation ( RAB7A , RAB7B , and the HOPS complex subunits VPS11 , VPS33A , VPS39 and VPS41 ) or addition of U18666A , which blocks late endosome-to-lysosome trafficking , were shown to inhibit both infection and virus-cell fusion . The importance of lysosomes for entry was confirmed by using knockout cells lacking a functional HOPS complex ( For a review on the HOPS complex see [54] ) . Interestingly , in these cells lysosomes are clustered in a perinuclear region of the cell rather than dispersed throughout the cytoplasm . Complementation of the missing HOPS subunit restored the normal lysosome distribution and entry of MHV ( Figure S7 in Text S1 ) . The importance of late endosome-to-lysosome trafficking for efficient entry was also observed in murine cells ( Figure S11C in Text S1 ) and for MHV-S4 carrying the S protein of MHV-4 ( JHM; Figures S10 and S12 in Text S1 ) . Corroborating the importance of trafficking of MHV virions through the endocytic pathway , perturbation of endosome maturation by the addition of inhibitory agents , such as ammonium chloride , BafA1 , Chloroquine , and Monensin inhibited infection and fusion of MHV . Also the importance of the actin and microtubule cytoskeleton - as demonstrated by the inhibition of MHV entry by downregulation of the Arp2/3 complex factors ( ACTR2 and ACTR3 ) , of the microtubule-associated transporter dynein ( DYNC1H1 and DYNC2H1 ) , or by addition of actin- or microtubule-affecting drugs - may be explained by the documented involvement of the cytoskeleton in endosome maturation ( reviewed in [7] ) . Indeed , entry of MHV-S2′FCS , which presumably fuses in early endosomes , was much less affected by actin-affecting drugs than that of MHV carrying wild type spike proteins ( Figure S10 in Text S1 ) . However , we cannot exclude that actin also plays a role in the clathrin-mediated uptake of MHV particles , as has been observed for VSV and other pathogens that depend on clathrin-mediated endocytosis ( reviewed in [91] ) . MHV particles require trafficking to the low pH environment of lysosomes to achieve membrane fusion . Nevertheless , MHV was much less sensitive to elevation of pH in the endo-/lysosomal system by the addition of BafA1 than viruses known to fuse in early or late endosomes ( VSV and IAV ) . BafA1 , an inhibitor of vacuolar-type H+-ATPase was effective in blocking MHV entry only at high concentrations , which are known to prevent endosomal maturation in addition to the elevation of the pH [66] . The absence of a functional HOPS complex , which is required for late endosome-to-lysosome maturation , did not affect infection of cells with VSV or IAV , while entry of MHV was severely reduced . Thus , the low pH trigger that mediates entry of VSV and IAV in the endosomal system of these cells , is not sufficient to induce fusion of MHV . Other environmental cues , present in lysosomes only , are apparently required to activate conformational changes in the S protein leading to fusion . Indeed , inhibition of the three major classes of proteases present in the lysosome by CPI effectively prevented MHV fusion . Infection of murine LR7 cells with MHV was also inhibited by CPI ( Figure S11B in Text S1 ) . Strikingly , other inhibitors that affect members of a single protease family had none or only little impact on MHV fusion . These results are in consistence with a functional redundancy of protease family members [47] , [76] and may explain why previous studies using specific lysosome protease inhibitors [27] , [92] failed to detect entry inhibition . Also , the inhibition of MHV entry by MG132 may be explained by the known ability of the proteasome inhibitor to negatively affect lysosomal proteases [93]–[95] , although we cannot exclude that MG132 affects entry by its interference with lysosomal trafficking [96] . Our results indicate that cleavage of the S protein immediately upstream of the FP is essential for CoV entry and determines the intracellular site of fusion . Although we did not demonstrate cleavage of MHV S at the FP proximal position directly , a recent study found a cleaved form of the MHV S2 subunit to correspond with the fusion-active form [49] . Furthermore , introduction of an optimal FCS at the FP proximal position abolished the entry inhibition by the pan-lysosomal protease inhibitor whilst introducing a dependency on furin-related enzymes . Consistent with the known presence of active furin in early endosomes ( reviewed in [78] ) the mutant virus no longer required trafficking to late endosomes/lysosomes for entry to occur . However , in the presence of furin inhibitor , entry of this mutant MHV was much more efficient in wild type cells than in cells lacking a functional HOPS complex ( Fig . 8 ) , indicating that under certain circumstances lysosomal proteases may play a role in entry of this virus as well . Trafficking of virions to lysosomes was shown to be also important for entry of FIPV , but not of MERS-CoV , in agreement with the latter virus containing a putative FCS immediately upstream of the FP . Correspondingly , entry of FIPV was inhibited by the pan-lysosomal protease inhibitor CPI but not by furin inhibitor , while the reciprocal held true for MERS-CoV . The importance of S protein cleavage downstream of the S1/S2 boundary and upstream of the FP for infection has so far only been demonstrated for SARS-CoV and IBV [40] , [43] , [46]–[48] . Based on the present study and on the work of others , we conclude that cleavage at the FP proximal position is likely to be a general requirement for CoV entry . With the exception of possibly IBV , cleavage at this position does not appear to occur in the virion-producing cell as it is not observed in released virions , but in the target cell ( this study; [40] , [43] , [47] , [48] ) . This suggests that receptor binding or other environmental cues are necessary to render the cleavage site accessible for proteolysis in the intact virion . Also for several other viruses , including RSV [13] and Ebola virus [16] , cleavage of the fusion protein upon endocytosis has been shown to be required for entry . Our results furthermore show that cleavage at a FP-proximal position is an important determinant of the intracellular site of fusion . The question remains , however , why some CoVs evolved to fuse in early endosomal vesicles while others require trafficking to lysosomes . In view of the growing number of proteases that have been shown to cleave CoV spike proteins [97] , this question should probably be studied in relation to the proteolytic enzymes available in the CoV target tissues and cells in vivo .
Murine LR7 fibroblast [98] and feline FCWF cells ( ATCC ) were used to propagate the recombinant MHV and FIPV viruses , respectively . HEK293T , MDCK and Vero cells were used to propagate pseudotyped VSVΔG/Luc-G* , Renilla luciferase expressing influenza A pseudovirus , or MERS-CoV , respectively , as described previously [71] , [73] , [99] . Cells were maintained as monolayer cultures in Dulbecco's modified Eagle's medium ( DMEM , Lonza ) , supplemented with 10% fetal bovine serum ( FBS ) . HeLa-ATCC cells stably expressing murine CEACAM1a ( HeLa-mCC1a ) and LR7 cells were used for infection experiments with MHV . HeLa-mCC1a cells stably expressing the deficient β-galactosidase ΔM15 ( HeLa-mCC1a-ΔM15 ) were used in the fusion assay . Stable cell lines were generated using a Moloney murine leukemia ( MLV ) retroviral vector . MLV was produced in HEK293T cells by triple plasmid transfection of a transfer vector containing the ΔM15 or mCC1a gene as well as a puromycin or neomycin resistance marker gene , respectively , in combination with expression vectors encoding the MLV Gag-Pol , and the VSV spike protein G . Upon MLV transduction , stably transduced cells were selected at 2 µg/ml puromycin and/or 0 . 5 mg/ml G418 ( both Sigma ) , maintenance at 1 µg/ml puromycin and/or 0 . 5 mg/ml G418 in DMEM , supplemented with 10% FBS . HAP1 cells and the VPS33A knockout derivative thereof ( H1-ΔV33 ) have been described previously [74] . H1-ΔV33 cells were stably transfected with FLAG-tagged VPS33A ( H1-ΔV33-fV33 ) using MLV transduction as described above using a blasticidin resistance marker gene in the transfer vector . Stably transduced cells were selected and maintained at 5 µg/ml blasticidin . HAP1 cells and its derivatives were also provided with mCC1 as described above to allow infection of these cells with MHV . The MHV fusion inhibitor HR2 peptide has been described before [100] and was synthesized by GenScript . The peptide was diluted in Tris/HCl 50 mM , pH 7 . 8 , 4 µM EGTA at 1 mM stock solution and used at 10 µM final concentration . Fluorescein-di-β-D-galactopyranoside ( FDG ) ( AnaSpec ) was dissolved in DMSO resulting in a 20 mM stock solution . Stocks of 700 mM cycloheximide ( CHX , Sigma ) , 125 µM Bafilomycin A1 ( BafA1 , Enzo Life Sciences ) , 140 mM Chloroquine ( Chloq , Sigma ) , 120 mM Dynasore ( Dyn , Enzo Life Sciences ) , 15 mM Dyngo-4a ( Dyngo , Abcam ) , 100 mM Ethylisopropyl amiloride ( EIPA , Enzo Life Sciences ) , 1 mM Nocodazole ( Noc , Sigma ) , 1 mM Latrunculin A ( LatA , Enzo Life Sciences ) , 2 mM Jasplakinolide ( Jasp , Sigma ) , 20 mM Cytochalasin B ( CytoB , Sigma ) , 20 mM Cytochalasin D ( CytoD , Sigma ) , 25 mM MG132 ( Sigma ) , 1 mM Brefeldin A ( BrefA , Sigma ) , and 10 mM Furin Inhibitor I ( FI , Calbiochem ) were prepared in DMSO and diluted 1∶1000 in the experiments , except when indicated otherwise . Stocks of 2 M ammonium chloride ( NH4Cl , Fluka ) , 5 mM AEBSF , 5 mM Leupeptin , 1 mM Camostat , 1 mg/ml Aprotinin ( all obtained from Sigma ) were prepared in H2O and used at 1∶100 final concentrations . 10 mM chlorpromazine ( Chlopro , Sigma ) , and 20 mM U18666A ( Enzo Life Sciences ) were prepared in H2O and used at 1∶1000 final concentrations . Stocks of 6 mM Monensin ( Mon , Sigma ) and 5 mM Phosphoramidon ( Sigma ) were prepared in methanol ( MeOH ) and used at 1∶1000 and 1∶100 final concentrations , respectively . 25 mg/ml cycloheximide ( CHX , Sigma ) and 5 mM Pepstatin A ( Sigma ) were prepared in methanol ( EtOH ) and used at 1∶1000 and 1∶100 final concentrations , respectively . Solvents EtOH , MeOH , and DMSO were obtained from Sigma-Aldrich . A stock of 125 µM CPI in PBS was made [76] and used at 5 µM final concentration . All plasmids were constructed using conventional cloning techniques . The ΔM15 gene was isolated from a DH5 E . coli strain by DNA extraction and PCR . The gene was cloned into a pCAGGS vector for ( transient ) expression and into a MLV-based pQCXIP transfer vector ( Clontech ) , resulting pQCXIP-ΔM15 , for the generation of stable cell lines . The gene encoding the MHV receptor mCC1a [101] was cloned into pQCXIN , resulting in pQCXIN-mCC1a . The RNA transcription vectors used for the generation of recombinant MHV using targeted recombination were generated using pMH54 derivatives [98] , [102] . pMH54 containing a GFP expression cassette between the E and M gene was generated as described previously for firefly luciferase [59] . The transcription vector used to generate MHV-S2′FCS ( pXHERLM-S2′FCS+ ) was generated by site-directed mutagenesis , thereby changing the sequence encoding AIRGR immediately upstream of the FP into a RRRRR-encoding sequence in vector pXHERLM [59] ( GCA′ATC′CGA′GGG′CGT to AGA′CGC′CGA′AGG′CGT ) . The transcription vector used to generate MHV-S4 expressing firefly luciferase , was generated by introducing the firefly luciferase expression cassette between the E and M genes similarly as described previously [59] in a pMH54-derived transcription vector that contains the gene encoding the S protein of MHV-4 ( MHV-JHM ) [82] . This latter vector was kindly provided by Susan Weiss . Recombinant MHV-EGFPM virus , containing a GFP expression cassette between the E and the M gene , MHV-S2′FCS , containing a Renilla luciferase expression cassette between the E and the M gene and a FCS at the FP-proximal position , and MHV-S4 containing the spike gene of MHV-4 ( JHM ) and a luciferase expression cassette were generated by targeted RNA recombination as described before [98] . Briefly , donor RNA was generated from linearized pMH54-derived transfer vectors described above , and electroporated into FCWF cells infected with interspecies chimeric fMHV coronavirus ( an MHV-A59 derivative , in which the ectodomain of the spike protein has been replaced by that of a feline coronavirus , thereby changing host cell tropism ) . The electroporated FCWF cells were seeded onto a monolayer of LR7 cells . After 24 h of incubation at 37°C , culture supernatant containing progeny viruses was harvested . Genotypes of the recombinant viruses were confirmed after two rounds of plaque purifications . Passage 3 stocks were used in experiments . Generation of MHV-EFLM and MHV-ERLM , containing a firefly or Renilla luciferase expression cassette between the E and the M gene , and MHV-αN , containing a N protein tagged with the α-peptide , has been described before [63] , [103] . Construction of FIPV expressing Renilla luciferase was reported previously [79] . Recombinant VSVΔG/FLuc-G* pseudovirus was generated as described before [71] . Construction of IAV-WSN pseudovirus expressing Renilla luciferase has also been described previously [73] . Viruses were stored in culture medium , supplemented with 25 mM HEPES or upon sucrose cushion purification in TN buffer ( 10 mM Tris-Cl , pH 7 . 4 , 10 mM NaCl ) . 30 , 000 HeLa-mCC1a- ( ΔM15 ) cells were seeded one day prior to transfection in a 24-well dish . Using Oligofectamine ( Life Technologies ) reagent three independent , non-overlapping siRNAs ( pre-designed Silencer Select siRNAs from Ambion ) per gene were individually transfected into target cells according to the manufacturer's instructions . Transfection mix for one well contained 2 . 5 µl of 1 µM siRNA and 0 . 5 µl Oligofectamine in 50 µl OptiMEM ( Gibco ) . Transfection was done in 250 µl final volume of OptiMEM . 4 hours post transfection 125 µl of DMEM , 30% FBS were added . Cells were infected 72 hours post transfection . HeLa-mCC1a cells were subjected to siRNA-mediated gene knockdown as described above . At 72 hpi cells were harvested by trypsinization , single-cell suspension counted , and collected by centrifugation . Cellular RNA was extracted using the RNeasy Mini Kit ( Qiagen ) . mRNA levels of genes were analyzed by qRT-PCR using a custom designed pair of specific primers to the gene resulting in about 150 bp products . RNA levels were measured using the GoTaq 1-Step RT-qPCR system ( Promega ) according to the manufacturers' instructions on a LightCycler 480 ( Roche ) . Expression levels were corrected for cell number and viability as determined by the Wst-1 assay ( Roche ) . Cells were inoculated with MHV-EGFPM at MOI = 0 . 5 ( 15–20% infected cells ) in DMEM , 2% FBS , for 2 h at 37°C . The inoculum was replaced by warm DMEM , 10% FBS . At 8 hpi , cells infected with MHV-EGFPM were trypsinized and fixed in 4% formaldehyde solution in PBS . Cells were washed and taken up in FACS buffer ( 2% FBS , 0 . 05M EDTA , 0 . 2% NaN3 in PBS ) and GFP expression was quantified by FACS analysis on a FACS Calibur ( Benson Dickson ) using FlowJo software . Of each sample at least 10 , 000 cells were analyzed . HeLa , LR7 , or HAP1 cells were inoculated with luciferase expressing ( pseudo ) viruses ( MHV-EFLM , VSVΔG/FLuc-G* , IAV-RLuc , MHV-S2′FCS , or FIPV-RLuc , MHV-EFLM-S4 ( JHM ) ) at MOI = 0 . 2 , unless indicated otherwise , in DMEM or IMDM ( HAP1 ) , supplemented with 2% FBS at 37°C . At 2 hpi the inoculum was replaced by warm culture medium containing 10% FBS . Cells were lysed at 7 hpi ( MHV , VSV , and FIPV ) or 16 hpi ( IAV ) in passive lysis buffer ( Promega ) . Firefly luciferase expression was assessed using the firefly luciferase assay system from Promega or using a homemade system ( 50 mM tricine , 100 µM EDTA , 2 . 5 mM MgSO4 , 10 mM DTT , 1 . 25 mM ATP , 12 . 5 µM D-Luciferin ) . Renilla luciferase expression was assessed using the Renilla luciferase assay system ( Promega ) . Light emission was measured on a Centro LB 960 luminometer . When indicated cells were transfected with siRNAs prior to inoculation as described above . Luciferase expression levels ( in relative light units , RLU ) were corrected for cell number and viability as determined by the Wst-1 assay ( Roche ) . When indicated cells were treated with pharmacological inhibitors starting at 30 min prior to or 2 h post inoculation . Huh-7 cells were inoculated with MERS-CoV at a MOI of 0 . 1 in FBS-containing DMEM . 8 h post infection , cells were fixed in 4% formaldehyde in PBS . Cells were stained using rabbit anti-SARS-CoV nsp4 antibodies that are cross-reactive for MERS-CoV , according to a standard protocol using a FITC-conjugated swine-anti-rabbit antibody . Number of infected cells was determined by cell counts on a wide-field fluorescent microscope . The β-galactosidase complementation fusion assay was performed as described previously [63] . Briefly , cells were preloaded with FDG substrate by incubation of adherent target cells with 2 . 5% FBS , 100 mM FDG , 50% PBS at room temperature . After 3 min incubation an excess of 5% FBS in PBS was added , supernatant removed and replaced by growth medium . After a recovery period of 30 min at 37°C , cells were ( mock ) treated with the different inhibitors for 30 min . MHV-αN virus was bound to cells in DMEM with 2%FCS ( in the absence or presence of inhibitors ) at a MOI = 20 for 90 min at 4°C to synchronize infection , after which cells were shifted to 37°C for 2 h . Cells were trypsinized and transferred to Eppendorf tubes , washed and immediately analyzed by FACS . For experiments with protease inhibitors the cells were loaded with FDG by hypotonic shock after trypsination and collection of the cells . In this case , FDG loaded cells were incubated on ice for 14 h before being analyzed by FACS . MHV wt virus was grown on LR7 cells and purified over a 20% sucrose cushion in TN buffer by centrifugation at 110 , 000 rcf for 2 . 5 h . Supernatant was removed and pellet resuspended in 200 µl TN buffer overnight on ice . Concentrated virus solution was subjected to further purification on a Pfefferkorn gradient ( 10–20% , 25–50% , 50% cushion ) . After spinning for 1 h at 150 , 000 rcf a clear virus band was visible . The virus band was collected and diluted in TN buffer . The virus was pelleted by centrifugation at 110 , 000 rcf for 1 h and resuspended in 200 µl 0 . 1M sodium phosphate , 0 . 15M NaCl buffer pH 7 . 2 overnight on ice . The purified virus solution was labeled using DyLight NHS 488 ( Thermo Scientific ) according to the manufacturer's instructions . Infectivity of the labeled virus was confirmed by TCID50 analysis and qRT-PCR . HeLa-mCC1a cells were seeded into 8-well glass-bottom chambers to reach 60% confluency the next day . Plasmids encoding mRFP-tagged RAB5A or RAB7A , or dsRed-LAMP1 [104] were transfected into the cells one day after seeding using Lipofectamine 2000 ( Life Technologies ) according to the manufacturer's instructions . 24 h after transfection MHV-DyLight488 was bound to cells on ice at MOI = 20 for 1 . 5 h in DMEM , 2% FBS . The inoculum was removed and cells washed with cold PBS to remove unbound virus . Warm imaging medium ( DMEM without phenol red , 10% FCS ) containing 0 . 008% trypan blue ( Invitrogen ) was added to the cell chambers . The cell membrane impermeable trypan blue shifts the expression spectrum of cell surface bound particles rendering them undetectable in the 505–530 nm channel ( described in [65] ) . Different low to medium RFP expressing cells were imaged live at 37°C , 5% CO2 in 10 min time frames from 10 min post warming up to 70 min in 30 s intervals thereby acquiring z-stack images . Each slice was 0 . 30 µm in thickness , averaging 12–14 slices per stack . For recording a Zeiss Axio Observer Z1 inverse spinning-disk confocal microscope , equipped with full box stage incubation , including CO2 ( Pecon ) , argon-krypton and helium-neon laser , two Photometrics Evolve 512 back-illuminated electron-multiplying charge-coupled-device ( EM-CCD ) cameras , and 100×1 . 46NA Oil alpha Plan Apochromat objective was used . Fluorescence images were exported as . czi files ( Zeiss ) and subsequently imported into Fiji ( ImageJ , NIH ) . Upon import into Fiji , color channels were split and saved as 8-bit tagged image file format . Virus movements were manually tracked in x/y or z direction in the green channel using the MTrackJ plugin . Tracks were saved and subsequently loaded onto the red channel . For each virus spot the area underlying a circle of 0 . 213 µm2 was measured for its gray mean value . Viruses were considered colocalizing if the gray mean value reached 50% of the maximum . Subsequently red and blue color channels were merged , tracks imported and viruses classified using the viral track . If the virus co-localized with the endosomal vesicle over at least four sequential 30 s frames the virus was categorized as associating . Viruses that , after initial co-localization , separated from the vesicle were classified as ‘associating/dissociating’ . If a virus particle faded and disappeared ( and could not be found in other z-stacks ) whilst co-localizing in previous intervals with an endosomal vesicle it was categorized as ‘fusing’ ( Figure S2 and S3 in Text S1 ) . When a viral particle co-localized with endosomal compartments but did neither dissociate nor fade during the 10 min acquisition period it was classified as ‘non-fusing’ . With this quantification method we analyzed 12 cells for RAB5 with 75 virions in total , 12 cells for RAB7 with 105 virions in total , and 16 cells for LAMP1 with 115 virions in total , acquired over three independent experiments . The sequences of MHV-A59 and MHV-S2′FCS were based on pMH54 sequencing results . Sequences for BCoV ( GI: 18033975 ) , FIPV ( GI: 556925469 ) , HCoV-OC43 ( GI: 530802591 ) , HCoV-HKU1 ( GI: 306569687 ) , SARS-CoV ( GI: 89474484 ) , MERS-CoV ( GI: 510937295 ) , HCoV-229E ( GI: 82780499 ) , HCoV-NL63 ( GI: 530802144 ) , IBV-Beaudette ( GI: 138186 ) were obtained from NCBI . Alignments were performed over the entire length of the spike proteins using MegAlign ( Lasergene DNASTAR ) using a ClustalW alignment , gap penalty 10 , gap length penalty 0 . 2 , delay divergent sequences 30% , DNA translation weight 0 . 5 , protein weight matrix: PAM series , DNA weight matrix: ClustalW . HeLa cells were co-transfected with mRFP-tagged RAB7A similarly as described previously [60] . Briefly , 7'500 HeLa cells were seeded one day prior to transfection in a 96-well plate . Using Oligofectamine ( Life Technologies ) reagent three independent , non-overlapping RAB7A siRNAs ( pre-designed Silencer Select siRNAs from Ambion ) per gene were individually transfected into target cells with the mRFP-RAB7A plasmid . Transfection mix for one well contained 2 . 5 µl of 1 µM siRNA , 10 ng plasmid , and 0 . 5 µl Oligofectamine in 12 . 5 µl OptiMEM ( Gibco ) . Transfection was done in 62 . 5 µl final volume of OptiMEM . 4 hours post transfection 125 µl of DMEM , 30% FBS were added . RFP expression was analyzed 24 h post transfection using an EVOS Cell Imaging System . Confluent HAP1 , H1-ΔV33 , and H1-ΔV33-fV33 cells and their stably mCeacam1a expressing counterparts were detached using a cell scraper , homogenized , and fixed . After 30 min incubation in blocking buffer ( 3% BSA ( Sigma ) , in PBS ) for 1 h cells were incubated in 1∶100 N-CEACAM-Fc [80] antibody , washed , and stained with 2ry AF488 goat-anti-rabbit antibody ( Life Technologies ) . After washing cells were analyzed by FACS at 10 , 000 gated single cells per sample . HAP1 cells were trypsinized and collected by centrifugation at 350 rcf for 10 min . The pellet was resuspended in Laemmli sample buffer containing 100 mM DTT , boiled for 5 min at 95°C and subjected to electrophoresis in 10% acrylamide ( Bio-Rad ) gels . Viruses were purified and concentrated over a 20% sucrose cushion ( in TN buffer ) at 110 , 000 rcf . Pelleted virus was resuspended in TN buffer overnight on ice . After addition of Laemmli sample buffer ( 1× final concentration , 100 mM DTT ) , samples were boiled for 5 min at 95°C and subjected to electrophoresis in 7% acrylamide ( Bio-Rad ) gels . Upon transfer to a nitrocellulose membrane ( Millipore ) , the presence of cellular and viral proteins was probed with antibodies against GM130 ( rabbit pAb , Abcam ) , FLAG ( HRP-labeled mouse anti-FLAG mAb , Sigma ) or the S2 subunit of MHV A59 [105] ( mouse anti-S2 mAb ) diluted 1∶1000 . When necessary , the blots were subsequently incubated with HRP-labeled rabbit anti-mouse or swine anti-rabbit antibodies ( both diluted 1∶5000; DAKO ) . Binding of HRP-labeled antibodies was visualized using Amersham ECL Western blotting substrate ( GE Healthcare Life Sciences ) according to the manufacturer's instructions . To image the localization of LAMP1 in HAP1 , H1-ΔV33 , and H1-ΔV33A-fV33 , cells the cells were seeded onto coverslips one day prior to staining . Cells were fixed in 4% formaldehyde in PBS for 15 min at RT , washed with PBS , and subsequently permeabilized in PBS containing 0 . 1% Triton-X-100 for 10 min . Cells were incubated with antibody against LAMP1 ( rabbit anti-LAMP1 pAb , 1∶100 dilution; Abcam ) in 3% BSA in PBS followed by incubation with secondary antibodies coupled to AF488 , AF-568 phalloidin , and DAPI ( all Life Technologies ) . The samples were analyzed using a confocal laser-scanning microscope ( Leica SPE-II ) . LR7 cells were infected at MOI = 0 . 1 or MOI = 4 . 0 with MHV-ERLM or MHV-S2′FCS in DMEM containing 2% FBS and 25 mM HEPES ( infection medium ) . After 3 h of infection supernatant was replaced by fresh infection medium and infection was allowed to progress over a period of 24 h . Every 3 h a small sample of the culture supernatant was collected and immediately frozen . The samples were subsequently analyzed in TCID50 assays on LR7 cells and subjected to qRT-PCR analysis to quantify virion production . Therefore viral RNA was extracted from the samples using the QIAamp Viral RNA Mini Kit ( Qiagen ) . The relative amount of viral RNA present was determined with a LightCycler 480 using LightCycler 480 RNA Master Hydrolysis kit ( Roche Applied Biosciences ) and specific primers and probe targeted against the MHV 1b gene by comparison with a standard curve . Gene SwissProt ID AAK1 Q2M2I8 ACTR2 P61160 ACTR3 P61158 CAV1 Q03135 CAV2 P51636 CLTC Q00610 DAB2 P98082 DNM1 Q05193 DNM2 P50570 DYNC1H1 Q14204 DYNC2H1 Q8NCM8 EPS15 P42566 FLOT1 O75955 FLOT2 Q14254 LAMP1 P11279 MYO6 Q9UM54 NSF P46459 PAK1 Q13153 RAB5A P20339 RAB5B P61020 RAB5C P51148 RAB7A P51149 RAB7B Q96AH8 SNX1 Q13596 VCL P18206 VPS11 Q9H270 VPS33A Q96AX1 VPS39 Q96JC1 VPS41 P49754 | Enveloped viruses need to fuse with a host cell membrane in order to deliver their genome into the host cell . In the present study we investigated the entry of coronaviruses ( CoVs ) . CoVs are important pathogens of animals and man with high zoonotic potential as demonstrated by the emergence of SARS- and MERS-CoVs . Previous studies resulted in apparently conflicting results with respect to CoV cell entry , particularly regarding the fusion-activating requirements of the CoV S protein . By combining cell-biological , infection , and fusion assays we demonstrated that murine hepatitis virus ( MHV ) , a prototypic member of the CoV family , enters cells via clathrin-mediated endocytosis . Moreover , although MHV does not depend on a low pH for fusion , the virus was shown to rely on trafficking to lysosomes for proteolytic cleavage of its spike ( S ) protein and membrane fusion to occur . Based on these results we predicted and subsequently demonstrated that MERS- and feline CoV require cleavage by different proteases and escape the endo/lysosomal system from different compartments . In conclusion , we elucidated the MHV entry pathway in detail and demonstrate that a proteolytic cleavage site in the S protein of different CoVs is an essential determinant of the intracellular site of fusion . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"viruses",
"viral",
"attachment",
"viral",
"entry",
"coronaviruses",
"feline",
"coronavirus",
"host",
"cells",
"rna",
"viruses",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"medica... | 2014 | Coronavirus Cell Entry Occurs through the Endo-/Lysosomal Pathway in a Proteolysis-Dependent Manner |
Dendritic cells ( DC ) , including those of the skin , act as sentinels for intruding microorganisms . In the epidermis , DC ( termed Langerhans cells , LC ) are sessile and screen their microenvironment through occasional movements of their dendrites . The spatio-temporal orchestration of antigen encounter by dermal DC ( DDC ) is not known . Since these cells are thought to be instrumental in the initiation of immune responses during infection , we investigated their behavior directly within their natural microenvironment using intravital two-photon microscopy . Surprisingly , we found that , under homeostatic conditions , DDC were highly motile , continuously crawling through the interstitial space in a Gαi protein-coupled receptor–dependent manner . However , within minutes after intradermal delivery of the protozoan parasite Leishmania major , DDC became immobile and incorporated multiple parasites into cytosolic vacuoles . Parasite uptake occurred through the extension of long , highly dynamic pseudopods capable of tracking and engulfing parasites . This was then followed by rapid dendrite retraction towards the cell body . DDC were proficient at discriminating between parasites and inert particles , and parasite uptake was independent of the presence of neutrophils . Together , our study has visualized the dynamics and microenvironmental context of parasite encounter by an innate immune cell subset during the initiation of the immune response . Our results uncover a unique migratory tissue surveillance program of DDC that ensures the rapid detection of pathogens .
The skin is the interface between the environment and internal tissues . Dendritic cells ( DC ) , as part of the body's innate immune defense , are strategically positioned in this organ; the epidermis is the home of Langerhans cells ( LC ) , while the dermis harbors dermal DC ( DDC ) . The main function of DC is believed to be the recognition and processing of foreign antigens , and subsequent presentation to naïve T cells [1] . DC normally reside in an immature state in the skin . Upon antigen encounter in the presence of “danger signals” , such as proinflammatory cytokines , DC undergo maturation enabling their migration to draining lymph nodes ( LN ) [2] . Accumulating evidence suggests that DDC may be responsible for the transport of pathogens to draining LN [3]–[6]; in certain infections , for example with herpes simplex virus , DDC act as an antigen shuttle , i . e . they transfer antigen to LN-resident CD8+ DC , which subsequently present it to T cells [7] . In other infections , including those with Leishmania parasites , they may present antigen directly to T cells [6] . Using intravital confocal microscopy , LC in the skin were found to be immobile with occasional repetitive dendrite movement , termed dendrite surveillance extension and retraction cycling habitude ( dSEARCH ) [4] , [8] . In contrast to LC , very little is known about the migratory and interactive behavior of DDC . This is of significance , as during certain infections DDC may come into close contact with microorganisms , and it is unclear whether DDC are capable of detecting living pathogens directly or take up antigens from dying infected cells or dead pathogens . Since these initial events of an immune response are likely to determine the magnitude and quality of T cell and B cell immunity , it is important to decipher the events of pathogen encounter directly in situ . Cutaneous Leishmaniasis is a disease caused by a large group of protozoan parasites belonging to the Genus Leishmania , including L . major . It serves as a paradigmatic skin infection , as promastigote stage parasites are directly deposited into the dermis during sand fly bites [9] . While it is thought that the parasites then infect innate immune cells in the skin , primarily macrophages [10] , [11] , the precise events occurring at the time of infection are not well defined . After entering cells , the parasites rapidly transform to the amastigote form , a rounded non-flagellated stage , which survives and multiplies within the phagolysosome ( parasitophorous vacuole , PV ) up until the time of cell rupture . After several weeks a lesion at the site of infection develops that is primarily composed of infected and inflammatory cells [12] , [13] . In some cases , these lesions are able to resolve over several months , while in others the lesions are chronic and can be associated with severe disease [9] . Current treatment options are scarce , therefore begging for the development of prophylactic vaccines . A prerequisite for this is a thorough understanding of the immune response against the parasites . Several studies have investigated the response of cutaneous DC to Leishmania spp . Initial reports suggested that LC are infectable by Leishmania , migrate to LN and activate CD4+ T cells [14] . However , more recently these findings have been questioned as DC harboring parasites in LN do not express the LC marker langerin [6] . Also , mice that lack MHC class II expression in LC but not DDC resolve infection similarly to wildtype animals [15] . While several investigators have suggested that DDC transport Leishmania to the paracortex of draining LN [5] , [6] , others have questioned the role of cutaneous DC during early infection altogether [16] , [17] . At later stages of disease , monocyte-derived DC may differentiate directly within the inflamed skin , and then migrate to draining LN where they induce CD4+ T cell activation [18] . To gain further insight into this controversy , i . e . what is the nature of parasite-DC encounter during early infection , ideally , Leishmania infections should be visualized directly in the natural microenvironment of the skin . In the present study , we made use of intravital two-photon microscopy ( 2P-IVM ) to address the following questions: 1 . What is the steady-state behavior of DDC ? 2 . How do DDC respond to danger signals ? And 3 . Do cutaneous DC take up Leishmania parasites in the early phase of infection , and if so , what are the dynamics of this process ? Surprisingly , we found that DDC were migratory under homeostatic conditions , which is in stark contrast to their epithelial counterparts . After encountering danger signals , DDC underwent a morphological transition into immobile , dendritic-shaped cells . At this point , the cells were capable of taking up parasites through the elaboration of motile pseudopods . Together , these results shed new light on the dynamics and anatomy of host-pathogen interactions .
In order to visualize the behavior of LC and DDC , we made use of CD11c-YFP mice [19] , in which DC express high levels of cytoplasmic YFP . To ascertain that skin DC expressed YFP , we analyzed single cell suspensions prepared from separated epidermis and dermis by flow cytometry ( Figure 1 ) . CD45+YFP+ epidermal cells were CD11c+CD11b+F4/80+I-Ab+ ( Figure 1 ) , and immunofluorescence staining of tissue sections showed that langerin expressing YFP+ cells displayed the characteristic morphology of LC ( data not shown ) . In the dermis , CD45+YFP+ cells were CD11c+CD11b+F4/80+I-Ab-high , and therefore represented DDC [20] . We also detected a subset of CD45+YFPlow cells within the dermis . However , this signal was due to autofluorescence , rather than specific YFP expression , as a similar population of cells was also found in wildtype animals ( Figure S1A ) . These cells were CD11c−CD11b+F4/80+Moma-2+I-Ab-low thereby resembling dermal macrophages [20] . The fluorescence intensity of these cells was , on average , 50 times dimmer than the YFP signal from DDC . Since , under our 2P imaging conditions , we did not detect any signal in the dermis of wildtype animals ( Figure S1B ) , we concluded that LC and DDC in CD11c-YFP mice can be detected by means of specific YFP expression , while other hematopoietic cell subsets remain undetectable . The distribution of YFP+ DC populations was determined by 2P-IVM in the ear skin of CD11c-YFP mice . Vertical scans revealed the presence of YFP+ cells between 5–20 µm below the outermost epidermal layer ( Figure 2A–2C ) . These cells exhibited numerous , irregularly shaped dendrites , morphologically consistent with LC . The highest density of LC was found 15 µm underneath the stratum corneum ( Figure 2C ) . Below the epidermis , second harmonic generation ( SHG ) signals highlighted extracellular matrix ( ECM ) components [21] forming a dense , mesh-like network ( vertical depth of 20–60 µm from the outermost surface ) . Embedded in the lower part of this network , with the highest density between 20–40 µm below the basement membrane and reaching up to a depth of ∼100 µm , were scattered YFP+ cells , of markedly different morphology to LC , i . e . of rounder shape , with fewer , shorter dendrites ( Figure 2A–2C ) . The overall density of LC was approximately 3 times higher than that of DDC ( Figure 2C ) . Together , these results established that cutaneous skin DC populations could be imaged by means of 2P-IVM , and identifed two morphologically distinct cutaneous DC subsets in the different compartments of the skin . While epithelial DC populations in the skin and gut have been found to be sessile [19] , [22] , no information is available on DC behavior in the interstitial space within peripheral organs . Nevertheless , DC in the dermis are suspected to be involved in antigen transport from the skin to draining LN thereby regulating the initial phases of host-pathogen responses . We therefore asked whether DDC scanned their microenvironment in a similar fashion to epidermal LC . To this end , we conducted time-lapse 2P-IVM in ear skin of CD11c-YFP mice . When focusing on the epidermis , we found that LC were sessile ( mean velocity <2 µm/min ) , with their dendrites remaining almost completely immobile ( Figure 3A–3C and Video S1 ) . As described previously , we occasionally observed dSEARCH [4] , [8] ( Figure 3B and Video S2 ) . However , in contrast to LC , we discovered that DDC were actively crawling through the interstitial space of the dermis at a mean velocity of 3 . 7±0 . 3 µm/min ( mean±SEM ) ( Figure 3A–3C and Video S3 ) . Migrating cells exhibited a polarized morphology , often displaying lamellipodia at the leading edge and a trailing uropod-like structure ( Video S3 ) . Since our experiments were performed in non-inflamed ear tissue , these results suggest that continuous migration is a steady-state feature of interstitial cutaneous DC . It may further indicate that the unexpectedly high motility of DDC serves to screen the dermal extracellular space for intruding microorganisms/environmental noxae . We next sought to define signals involved in spontaneous migration of DDC . When we treated animals with pertussis toxin ( PTX ) , an inhibitor of Gαi protein-coupled receptors , the capability of DDC to translocate within the dermis significantly decreased ( reflected by a reduction of their displacement; Figure 4A and Videos S4 and S5 ) . The migratory velocity of DDC did not change after PTX treatment , because cells moved back and forth in the same place ( therefore , following cell-centroids resulted in measurable velocity; Figure 4A and Video S5 ) . We concluded that , while PTX does not interfere with the migratory machinery of DDC per se , DDC utilize chemo-attractant signals , most likely chemokines , for their migration through the interstitial space . Having defined the cellular activities of skin DC in the steady-state , we determined their behavior in the presence of danger signals implicated in DC activation [23] . CD11c-YFP mice were injected intravenously with LPS ( 50 µg ) , which mimics systemic bacterial infection [24] . Two to eight hours after LPS treatment , LC remained sessile within the epidermis , without evidence of lateral or vertical movement ( Figure 4B ) . In contrast , we observed dramatic changes of DDC behavior two to four hours after LPS administration . They exhibited significantly decreased migratory velocity ( 2 . 12±0 . 21 µm/min ) and displacement , with more than 50% immobile cells ( Figure 4B , Figure S2 and Video S6 ) . Six to eight hours post LPS injection DDC partially regained their mobility ( 70% motile cells; Figure 4B , Figure S2 , and Video S7 ) . Thus , upon encounter of danger signals , DDC change their behavior , which may facilitate recognition/uptake of pathogens . To further test this hypothesis , we used the protozoan parasite L . major as a model pathogen . During natural infection , promastigote stage Leishmania spp . are directly deposited into the dermis by sand flies . Previous in vitro studies have demonstrated that DC can be infected by Leishmania parasites [13] , [25] . We therefore speculated that DDC may recognize and interact with L . major upon introduction into the dermis . 1–2×105 DsRed2-tagged Leishmania ( LmjF-DsRed2 ) promastigotes were injected in a small volume ( 1 . 5 µl ) of saline solution into the superficial dermis . This allowed us to deposit parasites underneath the epidermis at a vertical depth of 25–60 µm while keeping mechanical tissue disruption as minimal as possible ( Figure 5A ) . Within 20 min of injection , DDC in the vicinity of parasites decreased their migratory speed and changed their shape to a more dendritic cell-like morphology characterized by the emergence of multiple dendritic processes ( Figure 5B and 5C ) . This was paralleled by the appearance of several intracellular vacuoles , each of them containing a single red parasite ( Figure 5C ) , which is consistent with the formation of PVs [26] , [27] . Interestingly , these vacuoles were mobile , i . e . appeared to move freely within the cytoplasm of the cells . Two to three hours after infection , the percentage of DDC harboring one or more parasite was ∼70% ( Figure 5C ) . Of note , LC morphology and behavior was unchanged after infection with L . major . Further , LC were not found to take up parasites , at least during the first six hours of infection ( data not shown ) . However , it should be pointed out that parasites were injected intradermally . Consequently , LC access to parasites may have been prevented by anatomical barriers , such as the epidermal basement membrane . To determine whether parasite uptake by DDC was specific for the Friedlin strain ( FV1 ) of L . major , or could be recapitulated with other L . major strains , we injected the LV39 strain under the same conditions as described above . As shown in Figure 5C , this led to ∼55% of DDC containing parasites . We therefore consider L . major uptake by DDC a general phenomenon . For most of our experiments we made use of stationary phase L . major promastigotes . Since these cultures may contain stages of different infectivity or even a few dead parasites , confirmatory experiments ( n = 3 ) using highly purified metacyclic [28] LmjF-DsRed2 parasites were conducted . These experiments confirmed uptake of parasites by YFP+ DDC into cytoplasmic vacuoles to the same extent as stationary phase parasites ( Figure S3 ) . Since our LPS experiments had shown that DDC markedly reduce their locomotion after exposure to a danger signal , we next assessed the migratory behavior of L . major-bearing DDC . As shown in Figure 5D , infected DDC significantly reduced their migratory velocities . To determine whether parasite uptake and migratory arrest were related phenomena , we also measured the migratory speed of non-infected DDC . We found that the latter revealed a similar reduction in their migratory velocities and displacement as compared to their infected counterparts . Collectively , these results show that DDC , by default , reduce their migration at sites of inflammation . We also conducted sham infection experiments using a red fluorescent dye , SNARF-1 , in order to exclude that the physical manipulation during intradermal injection by itself caused changes in DDC behavior . As shown in Figure 5E , there was no difference in DDC migration between SNARF-1 injected and uninjected skin attesting to the specificity of the infection experiments . The exact mode by which Leishmania infects cells in vivo is not known . It is thought that parasite uptake by phagocytes involves non-random promastigote attachment to the cell followed by engulfment [29] . However , only in vitro data on this process are currently available , and the cellular and molecular mechanisms remain poorly understood . Our intravital imaging experiments demonstrated that cytoplasmic DDC processes actively extended towards parasites ( Figure 6A and Videos S8 and S9 ) at an average speed of ∼2 . 5 µm/min and reaching up to 50 µm in length . We occasionally observed that dendrite extension was preceded by parasite contact with the cell membrane followed by engulfment along the long axis of the parasite ( Figure 6A and Videos S8 and S9 ) . After capturing parasites , dendrites often rapidly retracted towards the cell body , paralleled by the formation of an intracellular vacuole . These results establish that L . major parasites in the interstitial space were internalized in a free form by DDC during the early phase of infection . We next asked whether inhibition of Gαi protein-coupled receptor signaling interfered with parasite uptake by inoculating mice with LmjLV39-DsRed2 parasites two to three hours after systemic PTX treatment . Since after PTX application DDC did not translocate through the dermis , we imaged cells that co-localized with the parasite depots . We observed that the formation of pseudopods and parasite uptake was preserved in these non-displacing DDC ( Figure 6B ) . This indicates that parasite sensing was independent of Gαi protein-coupled receptors . Furthermore , these results show that migration and dendrite formation can be uncoupled at the molecular level . Phosphoglycans ( PG ) , in particular lipophosphoglycans ( LPG ) , are essential during the infectious cycle of Leishmania . For instance , PGs have been implicated in the adherence of parasites to the gut epithelium in the sand fly , the resistance to complement in the blood stream , and have been considered candidate molecules for the uptake by host cells [13] , [30] . PG-deficient parasites persist in vivo for months without causing disease , and are therefore considered potential attenuated anti-Leishmania vaccine candidates [31] . While in vitro studies have shown that macrophages can take up PG-deficient parasites , it is not known whether the target cell range is the same for PG-deficient and wildtype parasites in vivo . To gain further insight into the role of LPG in parasite interactions with DC in vivo , we made use of DsRed2-tagged L . major deficient in the LPG2-encoded Golgi GDP-mannose transporter . These parasites fail to synthesize surface and other secreted PG [32] . As shown in Figure 6B , lpg2KO-DsRed2 parasite uptake was similar to that of LmjLV39-DsRed2 control parasites . Therefore , PGs appear to be dispensable in the initial sensing of parasites by dendrites as well as in the binding of parasites to the cell membrane and subsequent internalization . DC can , in principle , internalize a large variety of particulate material [33] . Thus , we next determined whether parasite uptake was a specific phenomenon , or whether DDC indiscriminately incorporate particles introduced into the dermis . When inert fluorescent beads were injected intradermally , a minority ( ∼20% ) of DDC revealed intracellular beads at a low number ( usually 1 bead/cell ) two to four hours after injection ( Figure 7A , Table 1 , and Videos S10 and S11 ) . When counting the ratio between particles present in the immediate vicinity of DDC ( i . e . within half a cell diameter ) and intracellular particles , it was obvious that there was a clear preference of L . major uptake ( ratio 2 . 7 ) as compared to bead uptake ( ratio 39 . 5; Table 1 ) . In addition , we never observed dendrite formation after bead injection . The bead injection procedure did not result in significant changes in migratory velocity or shape change of the cells ( Figure 7B ) , indicating that the mechanical trauma induced by the inoculation was not sufficient to alter DDC behavior . However , it was conceivable that an inflammatory stimulus may have increased phagocytic activity of DDC . To test this further , we co-injected beads and parasites . Interestingly , there was no increase in bead incorporation , demonstrating that DDC are capable of selectively discriminating between L . major parasites and inert material . Finally , we determined whether L . major uptake was a primary feature of DDC or was facilitated by other innate immune cells present in early infection . In particular neutrophils have been shown to serve as vectors for Leishmania uptake by macrophages [34] . Depletion of these cells prior to intradermal injection of LmjF-DsRed2 did not change the number of DDC containing parasites as compared to controls ( Figure 7C ) . While these results do not exclude a role of neutrophils in the defense against this pathogen , they suggest that parasite phagocytosis by DDC is independent of these cells .
DC are considered gatekeepers in the defense against intruding pathogens . While DC responses to microbes have been studied in great detail in vitro and in cells isolated ex vivo , very little is known about their interactions in the context of intact tissues in real time . The present study aimed to visualize , in a dynamic manner , the behavior of DDC in normal skin and in response to a defined pathogen . Using 2P-IVM , we found that , under homeostatic conditions , DDC were actively crawling through the dermal interstitial space . Remarkably , upon introduction of the protozoan parasite L . major , DDC transformed into stationary , dendritic-shaped cells that were capable of rapid parasite uptake through flexible dendritic processes . Together , our findings define the microenvironmental context of DDC-pathogen encounter in the earliest phase of cutaneous immune responses . That DDC migrate in the steady-state was unexpected , as other DC populations , such as DC in the gut epithelium and the epidermis have been found to be immobile , or in the case of the T cell area , very slow moving [4] , [8] , [19] , [22] . It is likely that the specific cellular motility patterns adopted by these individual DC populations serve to optimize their functions in their respective microenvironments . For instance , epidermal LC are in continuous close contacts with surrounding keratinocytes . The paucity of extracellular space may not require , or may not allow , movement of the cells for their immunosurveillance function . Rather , soluble antigens percolating through the extracellular epidermal space or signals from neighboring keratinocytes and/or adjacent LC may be sensed by the communicating dendrite network in this environment . DC in the LN T cell zones are characterized by extensive motions of their dendrites , which may be important for sensing of antigens filtering through the conduit system of this organ , and for establishing contacts with naïve T cells [19] . As compared to the epidermis , DDC are localized within the much more extensive dermal space , which , at the same time , contains considerably lower densities of resident cells , primarily fibroblasts . Thus , while keeping in mind that other tissue resident cells were not visualized in our study , DDC appeared as isolated cells embedded within the network of dermal ECM fibers . They were also found to be morphologically distinct from LC , i . e . they did not exhibit dendrites under non-inflammatory conditions . Therefore , the observation of their continuous crawling indicates a fundamental difference in tissue screening as compared to LC as well as DC in the T cell areas of LN . Since signals from intercellular communication by DDC with other dermal cells may be less abundant than for LC in the epidermis or DC in LN T cell areas , spontaneous DDC migration guarantees access to every corner of this specific microenvironment regardless of the activation state or potential damage of other resident cells during infection . Consequently , this ensures the rapid detection of intruding microbes and the subsequent immediate response to danger signals . Morphologically , DDC appeared to migrate in an amoeboid fashion , similarly to what has been described for T cells in the extravascular space [35] , [36] . Thus , crawling DDC exhibited an anterior-posterior asymmetry reflected by the formation of lamellipodia and uropods . This may suggest that similar molecular cues responsible for interstitial T cell migration , for example surface receptors involved in communication with the environment as well as intracellular molecules , mediate DDC locomotion . We found that blocking of Gαi protein-coupled receptors with PTX significantly reduced the displacement of DDC , implying chemoattractant receptors , such as chemokines or lipid mediators , in this process . This is consistent with recent 2P-IVM studies demonstrating that PTX inhibited the migration of naïve T cells within the LN paracortex , and that CCR7 is , at least partially , involved in this process [37]–[39] . However , the T cell zone of LN contains the fibroblastic reticular cell ( FRC ) network , which provides the structural backbone of this particular microenvironment . Elegant imaging experiments by Germain's group have shown that the FRC network acts as a scaffold for migrating naïve T cells [40] . A similar cellular structure does not exist in the dermis , raising the question as to how migrating DDC orient themselves within the dermis . It is conceivable that interactions with the extracellular matrix , primarily collagen fibers , are responsible for this process . Indeed , high resolution imaging has shown the intimate contact between DDC and the ECM ( Figure 3 ) , and it is likely that chemoattractants are deposited along these structures . Future studies will address the role of specific adhesion receptors , such as integrins or the hyaluronan receptor CD44 , as well as specific chemoattractant receptors in these interactions . The fact that DDC seemed to survey the dermis made us wonder whether they were indeed capable of detecting microorganisms introduced into the dermis . We chose the protozoan parasite L . major as a model pathogen , which is ideal in this context because , firstly , the parasite is directly deposited in the dermis during natural infection by sand flies , and secondly , the parasite is of sufficient size to be detected by 2P microscopy both extra- and intracellularly . Furthermore , the innate and adaptive immune responses against Leishmania spp . have been characterized in great detail in the past , even though controversy exists as to whether DC themselves are infected by the parasite during early infection ( reviewed in [11] , [13] , [25] ) . While the exact number of parasites transferred during sand fly bites is not known , inoculation of as few as 100 metacyclic parasites is sufficient for establishing an infection [41] . Although we could observe parasite uptake by YFP+ DDC by injecting as few as 2×104 parasites ( data not shown ) , this was technically challenging as only very few parasites and DC could be visualized in situ when using such low numbers . Thus , for the experiments in the present paper , 1–2×105 parasites were used in order to obtain data for proper statistical analysis . It should further be pointed out that the use of small volumes ( in the range of 1–2 µl ) for intradermal injection was critical , as larger volumes ( particularly >5 µl ) resulted in the disruption of the local microenvironment . This was evidenced by a disturbance of ECM fibers due to excess fluid ( edema ) and a migratory decrease/arrest of DDC within the injected ear , even after injection of saline solution without an inflammatory stimulus ( data not shown ) . In contrast , using our injection protocol , we did not observe a migratory or morphological change of DDC imaged ∼50–200 µm away from the injection site under control conditions ( Figures 5E and 7B ) . This result bears consideration not only for imaging studies , but for any situation in which the function of DDC is studied . In our intradermal infection model , we found that the majority of DDC picked up one or more L . major parasites shortly after inoculation . This was consistent when using two independent L . major strains , supporting the hypothesis that DDC are indeed capable of detecting this protozoan parasite in vivo . Interestingly , after the introduction of parasites , DDC underwent a morphological transition into bona fide DC-shaped cells . Strikingly , parasites appeared to be taken up by long , motile pseudopods ( Videos S8 and S9 ) . In vitro infection models of macrophages demonstrated that parasites initially adhered to the cell membrane in a non-random orientation , i . e . preferentially with either the tip or the base of their flagellum [29] . Subsequently , the parasites were engulfed by pseudopods wrapping around the parasites ( “coiled phagocytosis” [42] ) . We found that dendrite extension was sometimes preceded by parasite contact with the cell membrane , while at other times no visible contact was obvious . However , the level of 2P-IVM resolution did not always allow for unequivocal visualization of the parasite flagellum . Therefore , it is conceivable that physical contact is the main trigger of DDC dendrite extension observed in the context of Leishmania infection . Recently it has been suggested that DDC are composed of two separate subpopulations , i . e . the major langerin− subset and a small langerin+ subset [43]–[45] . Langerin+ DDC are distinct from in-transit LC , and have been shown to be capable of inducing cutaneous hypersensitivity reactions independently from langerin− cells . However , these cells are very rare ( 2–10% of DDC ) , and it is unclear whether their functions are different to langerin− cells . Since in our experiments 55–70% of all DDC are infected by L . major , the vast majority will represent langerin− cells . Together with previous studies showing that langerin− DC in draining LN present Leishmania antigens to T cells , we therefore speculate that langerin− DDC are the major players in this scenario . Ablation of langerin+ DDC using genetic approaches will enable definitive answers to this question . What are the mechanisms of parasite recognition by dendrites ? In the intestine , subepithelial DC have been found to extend processes between epithelial cells into the gut lumen , often revealing a spherical shape ( “balloon bodies” ) [22] , [46] . While these processes were capable of capturing bacteria in the gut lumen in a passive manner , this appeared to be a rare event . Importantly , sampling of gut material was non-discriminatory , i . e . DC did not distinguish between inert beads and bacteria [22] . In our study , we found that inert material ( beads ) alone or co-injected with parasites was largely ignored by DDC . In addition , when we injected fluorescently-tagged Bacillus Calmette-Guérin ( BCG ) , we found that two to four hours after inoculation only ∼30% of DDC contained single internalized BCG , comparable to the results using beads ( Figure S4 and data not shown ) . Furthermore , under these conditions we did not observe the transition of DDC into highly dendritic cells , even when they contained bacteria ( Figure S4 and data not shown ) . Together , these results suggest that L . major induces a specific change in DDC in vivo ( i . e . pseudopod formation ) , and may indicate the involvement of specific surface receptor ( s ) in this process . Previous studies have shown that Fc receptors and complement receptors are involved in Leishmania uptake by phagocytic cells . However , the molecular cues recognized on parasites are not well understood . Our studies have shown that PGs are not involved in parasite uptake by DDC . The use of parasite strains deficient in a variety of other structural and metabolic genes may , in the future , identify the molecular requirements of parasites to be sensed by dendritic processes . Another key observation from this study was the rapid transformation of migratory DDC into sessile DDC after exposure to microbial products , such as LPS . In addition , both infected and uninfected DDC became non-migratory at sites of L . major injection , suggesting that the inflammatory environment induces the change in migratory behavior , rather than parasite uptake per se . This conceivably also reflects a switch in functionality of DDC , i . e . from surveillance to sampling/antigen uptake . Thus , by arresting DDC in close proximity to the site of infection , they form a network of sessile cells “primed” for uptake of microbes present at the site . That these states are indeed distinct from each other is further reflected by the fact that PTX treatment interfered with DDC translocation , but not with parasite uptake . These findings raise the question as to the fate of DDC infected early during parasite infection . We have noted that DDC loaded with Leishmania remained relatively sessile over a period of up to ∼6 hours post-inoculation ( unpublished observation ) . When imaging at later time points ( ∼20 hours post infection ) , there were numerous YFP+ cells present within the dermis . While these cells showed a similar non-migratory phenotype as cells at earlier timepoints , the number of parasite-containing DDC decreased ( unpublished observation ) . Nevertheless , parasites were still present in the dermis , presumably within other cells ( unpublished observation ) . This may suggest that infected DDC leave the dermis at this stage in order to migrate to draining LN , and that these cells may be replaced by newly immigrating DC or their precursors from the bloodstream . Indeed , previous studies have shown that infected DC arrive in draining LN around 24 hours after infection [5] . Alternatively , parasites within DDC in vivo may simply lose fluorescence over time possibly due to an inability to survive for prolonged periods of time within these cells . Future studies will address potential interactions of infected DDC with the lymphatic vasculature in the dermis , and how these interactions are regulated at the molecular level .
Anti-mouse CD11b , CD11c , CD45 . 2 , F4/80 , I-Ab ( all from BD Biosciences ) , Langerin ( Dendritics , Lyon , France ) and Moma-2 ( Abcam , Cambridge , UK ) antibodies were used for flow cytometry analysis of epidermal and dermal cell suspensions . CD11c-YFP mice [19] ( kind gift of Dr . Michel Nussenzweig ) on a C57BL/6 background ( 10 generations ) were bred in the animal facility of the Wistar Institute and the Centenary Institute . Animal experiments were performed with approval of the Institutional Animal Care and Use Committees at both institutions . To generate fluorescent protein expressing L . major parasites , the gene encoding the red fluorescent protein DsRed2 was PCR amplified from pDsRed2 ( Clontech ) with primers that added BamHI sites to both ends . The PCR product was cut with BamHI and ligated into BglII site of pIR1SAT yielding pIR1SAT-DsRed2 ( strain B4787 ) . After SwaI digestion , it was introduced into L . major strain Friedlin V1 ( MHOM/IL/80/Friedlin ) by electroporation [47] . DsRed2-expressing LV39 clone 5 ( Rho/SU/59/P ) and its LPG2-deficient derivative were generated by stable transfection of Swa I-cut pIR1-SAT-LUC-DsRed2 ( B5947 ) . This plasmid was obtained by ligating the NruI-SalI DsRed2 fragment from pIR1SAT-DsRed2 ( B4787 ) into SalI+NruI digested pIR1SAT-LUC ( B5037 ) . Clonal transfectants were obtained and screened for bright red fluorescence and virulence in mouse infections ( data not shown ) . One clone of each strain was selected for work here ( L . major FV1 SSU:IR1SAT-DsRED2 ( b ) , LmjF-DsRed2; LV39 SSU:LUC:DSRED2 , LmjLV39-DsRed2; LPG2KO SSU:LUC:DSRED2 , lpg2KO-DsRed2 ) . Promastigotes were grown in complete M199 as described previously [47] . Red fluorescent protein expressing BCG was generated by transforming BCG Pasteur with plasmid pSMT3:mCherry ( a kind gift of Dr Wilbert Bitter , VU University Medical Center , Amsterdam , the Netherlands ) as previously described [48] . Hygromycin-resistant colonies were selected on Middlebrook 7H11 medium ( Difco Laboratories , Detroit , MI , USA ) and expanded in liquid Middlebrook 7H9 medium . Fluorescent colonies were selected by flow cytometry . Epidermal and dermal cell suspensions were prepared as described previously [49] with some modifications . In brief , ear tissues were incubated in trypsin ( 0 . 5% ) in HBSS buffer ( Invitrogen ) for 1 h at 37°C . For CD11c staining , we made use of dispase ( 5 U/ml ) instead of trypsin . After enzyme incubation , epidermis was separated from dermis . To obtain single cell suspensions , epidermal sheets were passed through a wire mesh , and dermal sheets were further digested with collagenase D for 1 hour . For Gαi protein-coupled receptor inhibition experiments , CD11c-YFP mice were injected intravenously with PTX ( 30 ng/g bodyweight ) in saline solution . For LPS experiments , CD11c-YFP mice were injected intravenously with 50 µg of LPS . 2P-IVM was performed at various time points after the injections . For neutrophil depletion , CD11c-YFP mice were injected i . p . with 250 µg of anti-Gr-1 antibody or rat IgG as control 24 hours before the inoculation of L . major promastigotes . Splenocytes from these mice were examined by flow cytometry to ensure the depletion of neutrophils at the end of imaging ( data not shown ) . In order to reduce autofluorescence from hairs , hair was removed from the ears for all imaging experiments [50] . Control experiments without the use of hair remover revealed identical migratory behavior of DDC ( Figure S5 ) . Mice were anesthetized by intraperitoneal injection of Ketamine/Xylazine ( 80/10 mg/kg ) . 1–2×105 stationary phase promastigotes in 1 . 5 µl of saline solution were injected intradermally using a 33 gauge Hamilton syringe . This procedure was performed under a stereoscopic microscope . For the bead experiments either 2 . 5×105 FluoSphere microspheres ( 2 µm , Invitrogen ) or 2 . 5×105 microspheres plus 2 . 5×105 FVI LmjF parasites were injected . For the BCG experiments , ∼2×105 BCG were injected intradermally . As an additional control , the fluorescent dye SNARF-1 ( 10 µg/ml ) was injected intradermally . Images were typically acquired 50–200 µm from the injection site . Anesthetized mice were placed onto a custom-built stage to position the ear on a small metal platform for 2P imaging . The ear was immersed in saline/glycerol ( 70∶30 , vol∶vol ) and covered with a coverslip . The temperature of the platform was maintained at 36°C , while the body temperature was regulated at 37°C through a heating pad placed underneath the mouse . Two-photon imaging was performed on a Prairie Technology Ultima System or a LaVision Biotec TrimScope equipped with a 40× ( NA 0 . 8 ) water immersion objective [35] . Both setups included four external non-descanned dual-channel reflection/fluorescence detectors , and a diode pumped , wideband mode-locked Ti:Sapphire femtosecond laser ( Coherent Chameleon or Spectra-Physics Mai Tai HP ) . The ear skin was exposed to polarized laser light at a wavelength of 950–960 nm . Three-dimensional ( x , y , z ) images of the ear skin were acquired ( 2–5 µm spacing in z-axis over a total distance of 10–25 µm ) every 30 s for a total observation period of 1–2 hours . Images acquired were then transformed into time sequence movies using Volocity software ( Improvision ) . Mean migration velocities , cellular displacement , and confinement ratios ( total length of track divided by distance between starting and end point ) were manually tracked and calculated for 15′ or 30′30″ as described previously [35] . Measurements were typically performed on 31 or 62 consecutive frames of the video . Cells were defined as immobile if the mean velocity was less than 2 µm/min [19] . To quantify the number of DC with internalized parasites , beads or BCG , images from 3D reconstructions of inoculated skin were examined for the colocalization of red signals ( L . major , beads or BCG ) and yellow signals ( DC ) . For comparisons , the Student's t test ( normally distributed ) or the Mann-Whitney test ( not normally distributed ) or one-way ANOVA were used . A difference was considered significant if P<0 . 05 . | Cutaneous Leishmaniasis is a difficult-to-treat disease affecting millions of people worldwide . Hence , there is high demand for the development of vaccines against Leishmania parasites , begging for a better understanding of immune responses against this pathogen . Dendritic cells , as part of the innate immune system , are thought to act as gatekeepers against intruding pathogens . However , their behavior in the context of intact tissues is incompletely understood . Here , we have used intravital two-photon microscopy to visualize the behavior of skin resident dendritic cells in real time , both in the steady-state and upon parasite encounter . We have found that migratory dermal dendritic cells are capable of rapidly sensing Leishmania parasites injected into the skin . This occurred through the formation of highly motile cellular processes capable of engulfing parasites , followed by parasite uptake into the cell . Together , our study provides a new vista of the orchestration of host cell–pathogen encounter in the three-dimensional context of intact tissues . Our results serve as the basis for a better understanding of the dynamic regulation of tissue surveillance by dendritic cells . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/protozoal",
"infections",
"immunology/immunity",
"to",
"infections"
] | 2008 | Migratory Dermal Dendritic Cells Act as Rapid Sensors of Protozoan Parasites |
Two diametric paradigms have been proposed to model the molecular co-evolution of microbial mutualists and their eukaryotic hosts . In one , mutualist and host exhibit an antagonistic arms race and each partner evolves rapidly to maximize their own fitness from the interaction at potential expense of the other . In the opposing model , conflicts between mutualist and host are largely resolved and the interaction is characterized by evolutionary stasis . We tested these opposing frameworks in two lineages of mutualistic rhizobia , Sinorhizobium fredii and Bradyrhizobium japonicum . To examine genes demonstrably important for host-interactions we coupled the mining of genome sequences to a comprehensive functional screen for type III effector genes , which are necessary for many Gram-negative pathogens to infect their hosts . We demonstrate that the rhizobial type III effector genes exhibit a surprisingly high degree of conservation in content and sequence that is in contrast to those of a well characterized plant pathogenic species . This type III effector gene conservation is particularly striking in the context of the relatively high genome-wide diversity of rhizobia . The evolution of rhizobial type III effectors is inconsistent with the molecular arms race paradigm . Instead , our results reveal that these loci are relatively static in rhizobial lineages and suggest that fitness conflicts between rhizobia mutualists and their host plants have been largely resolved .
Eukaryotes universally encounter bacteria that inhabit , infect , and often provide them with significant fitness benefits . In many cases , bacterial mutualist lineages exhibit intimate interactions with hosts , giving each partner opportunity to shape the phenotype of the other . Two diametric paradigms remain unresolved for the co-evolution of bacterial mutualists with their eukaryotic hosts [1] . One common paradigm models mutualist-host interactions as an antagonistic arms race , as is the case for co-evolution of pathogens and their hosts . Under this model , natural selection is predicted to shape partners to rapidly evolve traits to maximize their own selfish gains from the interaction and minimize costs invoked by the other [1] . This paradigm predicts that there is constant conflict over the fitness gain that each partner receives from the interaction even though both partners can attain net fitness benefits . The alternative framework assumes that conflicts between microbe and host are largely resolved [2] , [3] . It is predicted that the common genotypes are more likely to find compatible partners than the rare genotypes . As a consequence , the interaction is predicted to exhibit evolutionary stability , with lower rates of evolutionary change . Testing these competing frameworks by comparing the genetic patterns of known host-association genes between mutualists and pathogens will help to examine whether bacteria-eukaryotic mutualisms represent reciprocally exploitative interactions , as they have often been characterized , or alternatively , if these interactions exhibit a “mutualistic environment” in which evolutionary stasis is maintained [1] , [3] . A striking and well-studied example of arms race co-evolution occurs between proteobacterial pathogens and plant hosts . Plants have multiple defense systems to recognize and respond to bacterial infection . One key plant defense is pattern-triggered immunity ( aka PAMP-triggered immunity; PTI ) , in which pattern recognition receptors detect conserved microbe-associated molecular patterns and trigger defenses [4] . To counteract host defenses , many phytopathogenic bacteria use type III secretion systems ( T3SS ) to deliver collections of type III effector proteins ( T3Es ) to dampen host defenses , thereby allowing the bacteria to proliferate within host tissues and cause disease . A second line of host defense is effector-triggered immunity ( ETI ) in which resistance ( R ) proteins surveil for corresponding microbial effectors to trigger a robust defense often associated with a localized programmed cell death ( hypersensitive response; HR ) . Plant pathogen T3Es exhibit patterns of genetic variation that reflect rapid evolution , as predicted by the antagonistic arms race model [1] , [5]–[8] . In Pseudomonas syringae , the phytopathogenic species with the most extensive experimentally-validated set of T3Es , strains vary dramatically in T3E gene content , both in terms of the total number and sequence of effector genes [1] , [7] . Even highly related strains exhibit T3E presence/absence polymorphisms and insertion/deletion mutations that affect their coding sequences [2] , [3] , [9] . An important aspect of pathogen T3E collections is that their robustness is ensured via T3E redundancy so that any individual T3E gene is dispensable [1] , [3] , [10] . Hence , under the arms race scenario , rapid evolution of T3Es is advantageous to phytopathogens as novel collections of T3Es are more likely to avoid recognition while balancing sufficiency in subverting host defenses . Functional T3SS orthologs have been uncovered in diverse mutualistic species of proteobacteria , including nitrogen-fixing rhizobial species Sinorhizobium fredii ( Ensifer fredii ) , Bradyrhizobium japonicum , and Mesorhizobium loti [4] , [11] , [12] . Analyses of T3SS and T3E ( Nops; Nodulation Outer Proteins ) of rhizobia reveal many parallels to those of phytopathogens , pointing to the possibility that rhizobial nop genes are also under selection to maximize rhizobial fitness at potential expense of the fitness of the host . For instance , multiple studies have shown that T3SS and Nops of rhizobia are necessary for the establishment of mutualist infections and can modulate host PTI [13]–[20] . Moreover , T3Es of rhizobia also risk detection by host defense surveillance systems . In fact , legume loci responsible for “nodulation restriction” are R genes that restrict rhizobia in a T3SS-dependent manner and are linked to loci associated with resistance against phytopathogens [21]–[24] . This is consistent with the repeated observations that rhizobial strains deleted of genes encoding T3SS-secreted proteins gain new hosts that were once incompatible [19] , [24] , [25] . Since no study has examined the molecular evolution of T3Es in the context of mutualism , it is presently unknown whether these lineages exhibit patterns of genetic variation that would reflect arms race evolution with their hosts [5]–[8] . To this end , we investigated the molecular evolution of T3E genes in two lineages of mutualistic rhizobia and tested the arms race versus mutualistic environment paradigms . We used an experimentally validated set and compared their genetic patterns against the patterns of T3Es from five monophyletic strains of P . syringae ( group I strains ) and four that infect legumes ( legume pathovars ) to test the null hypothesis that collections of T3Es of mutualists evolve in a manner similar to those in proteobacterial phytopathogens .
We selected three S . fredii and five B . japonicum strains based on the criterion of demonstrable reliance on T3SS for host infection [11] . For B . japonicum , we also chose strains based on the genetic diversity inferred from their phylogenetic relationship [26] . At the initiation of this study , the only available finished genome sequences were for S . fredii NGR234 and B . japonicum USDA110 [27] , [28] . We used paired-end Illumina sequencing to generate draft genome sequences for S . fredii USDA207 , USDA257 , and B . japonicum USDA6 , USDA122 , USDA123 , and USDA124 ( Table S1 ) . Initial and post hoc analyses based on comparisons to reference and corresponding finished genome sequences completed subsequent to our efforts , respectively , indicated that the assemblies and annotations are of sufficient quality and covered the majority of the genomes for whole-genome characterization and comprehensive genome mining ( Figure S1 ) . We next used multiple measures to compare the within-group diversity for the rhizobial groups to that of the group I and legume P . syringae pathovars to determine the suitability of the latter two for genomic comparisons ( Figure 1; [7] ) . Quantitative measures of phylogenetic diversity ( PD ) fell within a narrow range with the two rhizobial groups having the higher PD values [29] . We also compared bacterial group PD values to those derived from equally sized groups of strains randomly assigned from the 17 used in this study . The within-group diversity of S . fredii , B . japonicum , and P . syringae , are similar , marginally , and significantly lower , respectively , relative to expectations due to chance . Additional measures based on average reliable single nucleotide polymorphisms ( SNPs ) per kilobase ( kb ) and average percent of orthologous pairs of genes were also consistent ( Figures 1 , S1 , and S2 ) . In total , the data demonstrate that the levels of genome-wide , within-group genetic diversity are higher in the S . fredii and B . japonicum groups , respectively , relative to either of the P . syringae groups . Candidate T3E genes were identified based on their association with a tts-box , a cis element proposed to be recognized by TtsI , a regulator of T3SS genes in rhizobia [13] , [16] . We identified a total of 305 putative tts-boxes ( Table 1 ) . In S . fredii NGR234 , we identified two additional tts-box sequences that were not previously reported [13] . In the finished genome sequence of B . japonicum USDA110 , we found 52 tts-boxes , of which 29 were previously identified ( Table 1; [30] ) . Fourteen of these tts-boxes are located upstream of 13 genes ( bll1862 has two upstream tts-boxes ) that encode proteins that are secreted in a T3SS-dependent manner [30] . We searched up to 10 kb downstream of the 305 tts-boxes and identified a total of 268 candidate T3E genes that clustered into 92 different families ( Table 1 ) . We adopted the Δ79AvrRpt2 reporter in the γ-proteobacterium P . syringae pv . tomato DC3000 ( PtoDC3000 ) for high throughput testing of candidate rhizobial T3E for T3SS-dependent translocation into plant cells , the most important criterion for defining a T3E [31] . We first selected NopB and NopJ from S . fredii NGR234 as likely T3E candidates for validation of heterologous T3SS-dependent translocation . NopB is secreted in vitro in a flavonoid- and T3SS-dependent manner from S . fredii NGR234 , and NopJ is a member of the YopJ/HopZ T3E family [32] , [33] . PtoDC3000 carrying either the nopB::Δ79avrRpt2 or nopJ::Δ79avrRpt2 fusions elicited HRs within the same time frame ( ∼20 hours post inoculation; hpi ) and to the same degree as the positive control , a fusion between the full-length avrRpm1 P . syringae T3E gene and Δ79avrRpt2 ( Figure 2A ) . Although Arabidopsis ecotype Col-0 can elicit ETI in response to both AvrRpm1 and AvrRpt2 , the observed HR is known to be a consequence of perception of the latter by RPS2 [34] . Each of the tested nopB::Δ79avrRpt2 gene fusions were sufficient for PtoDC3000 to trigger an HR at 20hpi , confirming that this family encodes bona fide T3Es ( Figure 2B ) . The NopB family is polymorphic with NopBNGR234 sharing ≥98% amino acid identity with NopBUSDA207 , but only 32% with NopBUSDA110 . In contrast , PtoDC3000 lacking fusions to Δ79avrRpt2 failed to elicit an HR but eventually showed tissue collapse approximately 28 hpi , indicative of PtoDC3000-caused disease symptoms ( data not shown ) . The T3SS-deficient mutant of PtoDC3000 ( ΔhrcC ) , regardless of the gene it carried , failed to elicit any phenotype throughout the course of the study , thereby demonstrating the T3SS-dependent delivery of T3Es ( Figure 2A ) . The demonstration that members of a polymorphic T3E family behaved identically in the heterologous delivery assay allowed us to test just a subset of 127 genes that represent the diversity present in the 268 candidates . From these , 87 T3Es belonging to 47 families between the two rhizobial lineages were confirmed for T3SS-dependent translocation ( Table 1; Figure 3 ) . We also used the sequences of members of confirmed T3E families to re-survey all draft genome sequences and identified an additional 21 homologs that were interrupted by physical and sequence gaps . Nine CDSs were amplified using PCR and sequenced and all were classified as functional based on the absence of premature termination codons . The remaining 12 genes belonged to 10 families with four homologs having upstream sequences similar to a tts-box but with bit-scores below our threshold . Seven had no discernible upstream tts-box , and one ( nopM2 ) potentially represents a subgroup of the nopM family since two copies are present in B . japonicum USDA123 . Of the 24 candidate and confirmed T3E families that were identified prior to this study , our computational method identified 21 of which 19 were experimentally validated as T3Es ( Table S2; [35] , [36] ) . Of the five that we failed to confirm , NopA may in fact be a secreted structural component of the T3SS [37] . NopT , in contrast , is likely a bona fide T3E but its cytotoxic effects in Arabidopsis could have caused misleading conclusions in the translocation assay [18] . NopC , NopH and NopD lacked a detectable tts-box or failed to meet the requirement of being >100 amino acids in length . The T3Es were assigned to families according to guidelines developed for T3Es of pathogenic bacteria [38] . Newly identified T3E families were assigned NopY through NopBT whereas 16 previously named families , that were confirmed in this study as representing T3Es , remain unchanged [11] . A relational table of the validated T3E genes is provided ( Table S2 ) . Other than those previously identified , none of the translated sequences of the T3E genes identified in this study have detectable homology to proteins of known function . We compared the genetic patterns of the rhizobial T3Es to: 1 ) those of the group I strains of pathogenic P . syringae , and 2 ) the core genome of the respective bacterial groups ( i . e . , genes ubiquitous to all strains within a group ) to test the null hypothesis that rhizobial T3Es exhibit signatures of arms race evolution similar to what has been characterized in pathogenic P . syringae lineages [5]–[8] . The T3Es of rhizobia were predominantly core , unlike the T3Es of the group I strains of P . syringae ( Figure 4A ) . In fact , the representation of T3Es among the four categories of core , singletons ( present in only a single strain of a group ) , pseudogenes ( premature termination codon relative to a full-length family member ) , and other ( polymorphic in regards to presence/absence ) , was significantly different ( Figure 4B ) . Next , we compared the proportion of core and accessory T3E genes in the S . fredii , B . japonicum , and P . syringae group I strains to the proportion of genes that are core and accessory to each group ( Figure 4C ) . Analysis indicated that the proportions of core T3E genes were significantly more than core genes for both groups of rhizobia . In contrast , the proportion of core T3E genes for the group I strains of P . syringae was significantly less . Thus , the collections of T3E genes of rhizobia are significantly more conserved than the collection of T3E genes of P . syringae and relative to their core genomes . The sequences within T3E families of S . fredii and B . japonicum are also highly conserved , as more than 75% of the within-family pairwise comparisons had ≥90% amino acid identity ( Figure 5A ) . Strikingly , twenty of the T3E families had all members with ≥99% identity . The T3Es of the group I P . syringae strains have a wider distribution in amino acid identity and a greater number of presence/absence polymorphisms than S . fredii or B . japonicum . Even when the latter variation was excluded from analysis , S . fredii and B . japonicum exhibit significantly more amino acid conservation of T3Es than group I P . syringae , whereas there was only marginal difference between the rhizobial lineages ( Figure 5B ) . To determine whether the levels of sequence conservation of T3E gene families differed relative to genes core to their respective genomes , we calculated and compared the within-family amino acid identity for the translated sequences of gene families core to each of the groups ( Figure 5C ) . The T3E gene families were significantly more conserved in sequence in both groups of rhizobia . In contrast , for the five group I P . syringae strains the translated sequences of the T3Es exhibited significantly lower amino acid identities as compared to the translated sequences of the core gene families . Therefore , relative to their respective core genes , the T3E genes of rhizobial and P . syringae lineages differed , with the former displaying higher levels of sequence conservation and the latter having significantly lower conservation . There is no clear relationship between T3Es and host range of pathogens [7] , [39] . P . syringae strains that infect the same host possess substantially different collections of T3Es . For example , P . syringae pairs , pvs . tomato races DC3000 and T1 and lachrymans races 106 and 107 , share no more than 50% of their T3Es in common [7] , [40] . The high variability in T3Es is in spite of the lower levels of genetic diversity detected relative to most pairs of rhizobial strains ( Figures 1 and S2; [9] ) . It has been suggested that Xanthomonas pathovars with similar hosts share similar compositions of T3Es [41] , [42] . However , the genome-wide diversity is unknown for these bacteria and furthermore , use of contemporary methods to study two Xanthomonas species has revealed a surprisingly high number of pseudogenized T3Es and divergence in T3E collections [43] , [44] . For P . syringae , it is hypothesized that T3Es are capable of functioning in a range of plant species [39] . The extensive host ranges for two of the rhizobial strains studied herein support this notion . S . fredii NGR234 and USDA257 can infect 112 and 79 genera of host plants , respectively , many of which are not considered cultivated plants and are more apt to have high within-population genetic diversity [45] . Support is further bolstered by the observation that P . syringae deleted of a T3E gene gains the ability to infect an otherwise non-host plant [46] . Similarly , rhizobial mutants deficient in secretion of T3SS-associated proteins can gain new species of plants as hosts [19] , [24] , [25] . To further test the potential for host range as a factor in the conservation of rhizobial T3Es , we compared their genetic patterns to those of T3Es from four P . syringae pathovars that like rhizobia , can infect legumes as hosts [7] . First , we compared between the two P . syringae groups . As expected , the within-group genetic diversity is similar ( Figures 1 and S2 ) . The genetic patterns of T3Es of the legume pathovars do not deviate significantly from those of the group I strains ( Figures 1 , 4 , and S2 ) . Finally , relative to core genes , the core T3Es of the legume pathovars have genetic patterns that are significantly different ( Figure 4C and 5C ) . Thus , despite the fact that the legume pathovars are distributed between two P . syringae groups , they exhibit similar levels of genome-wide and T3E diversity as the group I strains . Relative to the T3Es of S . fredii and B . japonicum , the T3Es of the legume pathovars are significantly more variable . As was the case for comparisons to the group I strains , the representation of core , singletons , pseudogenes , and other T3E categories is significantly different between legume mutualist and legume pathovars ( Figure 4A and 4B ) . Likewise , the T3Es of the legume mutualists have a significantly different distribution in percent amino acid identity within T3E families relative to those of the legume pathovars ( Figures 5A and 5B ) . Therefore , we conclude that a difference in host range is not a likely explanation for the extreme contrasts in T3E conservation between mutualist and pathogen . The type III secretion system is a key mechanism used by a diversity of bacterial mutualists to establish infections with their hosts . We identified and validated type III effectors to test the two diametric frameworks of mutualist-host co-evolution ( Figures 2 and 3; Table S2 ) . Rhizobial T3E genes show genetic patterns indicative of surprising conservation , pointedly contrasting the patterns consistent with the dynamic arms race model of co-evolution dogmatic for T3Es ( Figures 3–5 ) . This finding is particularly striking in light of the observations that T3Es of mutualistic rhizobia are similar in regards to those of pathogens in having to maintain sufficiency in engaging and dampening PTI while avoiding ETI [15] , [17] , [24] , [47] . Moreover , we demonstrated that the high conservation of T3Es in rhizobia relative to phytopathogens is not likely driven by differences in host range or phylogenetic diversity among genomes ( Figures 1 , S2 , 4 , and 5 ) . The high conservation in sequence and the fact that most of the T3E loci are co-localized are also consistent with acquisition events by both species of rhizobia . In B . japonicum , for example , most of the T3E genes are found distributed throughout an ∼700 kb-long symbiosis island . However , analysis of B . japonicum USDA110 and USDA6 suggested that the symbiosis islands were acquired independently , arguing against a common genome innovation event [48] . We favor an alternative explanation that the relative conservation of rhizobial T3Es reflects the selective pressures in these beneficial plant-microbe interactions . It has been suggested that legume hosts exhibit less polymorphisms in loci that restrict nodulation , in contrast to the higher levels of polymorphisms observed in loci that mediate resistance against phytopathogens [49] . Our data support this idea that novelty in mutualism can result in instability , specifically that rhizobial mutualists may be under pressure by the host that limits diversification [50] , [51] . In this context , hosts select for the most beneficial rhizobial genotype and these consequently common genotypes are more likely to find a suitable host . The type III effectors of S . fredii and B . japonicum thus exhibit mutualistic co-evolution with host defenses .
Bacterial strains used in this study were: S . fredii strains USDA207 and USDA257; S . fredii ( aka Rhizobium sp . ) NGR234; B . japonicum strains USDA6 , USDA110 , USDA122 , USDA123 , and USDA124; PtoDC3000 , its T3SS-deficient mutant ( ΔhrcC ) , and Escherichia coli DH5α . Rhizobia strains and P . syringae were grown in modified arabinose gluconate media ( MAG ) or King's B ( KB ) media , respectively , at 28°C . E . coli DH5α was grown in Luria-Bertani ( LB ) media at 37°C . Antibiotics were used at the following concentrations: 50 µg/ml rifampicin ( PtoDC3000 ) , 30 µg/ml kanamycin ( all bacterial strains ) , 50 µg/ml chloramphenicol ( B . japonicum strains ) , and 25 µg/ml gentamycin ( E . coli ) . Genomic DNA was extracted from S . fredii strains USDA207 and USDA257 and B . japonicum strains USDA6 , USDA122 , USDA123 , and USDA124 using osmotic shock , followed by alkaline lysis and phenol-chloroform extraction . We prepared 5 µg of DNA from each strain according to the instructions provided by the manufacturer ( Illumina , San Diego , CA ) . Paired-end sequencing was done by the Center for Genome Research and Biocomputing Core Labs ( CGRB; Oregon State University , Corvallis , OR; Table S1 ) . Velvet 0 . 7 . 55 was used to de novo assemble paired-end short reads [52] . Multiple assemblies , using different parameters , were produced for each genome and the highest quality assembly was identified using methods described previously [53] . Genomes were annotated using Xbase and further refined using the NCBI conserved domain database ( CDD; [54]–[61] ) . The Mauve Aligner 2 . 3 ( default settings ) program was used to compare the draft and finished genomes and , in other instances , reorder contigs to reference sequences [62] . To identify SNPs , we used Bowtie ver . 0 . 12 . 5 to align short reads to the finished genome sequence , allowing up to two mismatches [63] . Reliable sequence differences were identified based on having coverage of ≥10 reads and ≥8 reads supporting the same alternative base call . For P . syringae , we treated the publicly available genome sequences as true and incremented along the genome in 1 base pair increments , shearing in silico the genome into 32mers , and aligned the sequences to the indicated reference genome sequence . Homologous sequences were identified using reciprocal BLASTP analysis ( e-value≤1×10−15; >50% length of sequence ) of translated sequences ( those <50 amino acids in length were excluded ) . The Circos plot was generated using the Circos Table Viewer [64] . Genome sequences were retrieved from http://www . ncbi . nlm . nih . gov/genome: S . fredii NGR234 ( NC_012587 ) , S . fredii USDA257 ( NC_018000 ) , B . japonicum USDA6T ( NC_017249 ) , B . japonicum USDA110 ( NC_004463 ) , the P . syringae pathovars , actinidiae ( Pan; AEAL00000000 ) , glycine ( Pgy R4; ADWY00000000 ) , lachrymans ( Pla 106; AEAM00000000 ) , morsprunorum ( Pmp; AEAE00000000 ) , phaseolicola ( Pph 1448a; NC_005773 ) , pisi ( Ppi R6; AEAI00000000 ) , syringae ( B728a; NC_007005 ) , tomato ( PtoDC3000; NC_004578 ) , and tomato ( Pto T1; ABSM00000000 ) . Finished genome sequences from S . fredii USDA257 and B . japonicum USDA6T were used for post hoc analysis of genome assemblies [48] , [65] . We used HAL ( default settings ) to identify clusters of orthologous genes and generate a whole-genome phylogeny of the 17 strains plus two δ-proteobacterial reference strains , Geobacter sulfurreducens PCA ( NC_002939 ) and Desulfovibrio vulgaris RCH1 ( NC_017310 ) , used as outgroups [66] . PD values were calculated using the Picante R package [67] . To calculate PD values for randomly assigned groups of three , four , five , and five strains , an ad hoc Perl script was used to randomly assign the 17 rhizobial and P . syringae strains into four groups . The process was iterated 1000 times and PD values were calculated for each group per iteration . Statistical significance was determined by comparing the observed PD values to the proportion of 1000 iterations that had higher or lower PD values than the observed PD values . To identify the proportion of core genes for each group of strains , we identified the clusters of orthologous genes , generated by HAL , that were represented by all strains within each group . Fisher's exact test was used to compare the representations of T3E genes in the four categories for all possible pairs of bacterial groups [68] . The Kolmogorov-Smirnov test was used to compare the distributions of percent amino acid identity of T3E genes for all pairwise comparisons [69] . We developed a linear regression model that evaluates the average percent amino acid identity for both core and T3E families , using the core genes in the group I strains of P . syringae as the baseline:where Y = the response variable , percent amino acid identity; P = 1 for the legume pathovars of P . syringae and P = 0 otherwise; B = 1 for the B . japonicum species and B = 0 otherwise; S = 1 for the S . fredii species and S = 0 otherwise; E = 1 for T3E families and E = 0 for core gene families; ε = random error . Specifically , ß0 measures the average percent amino acid identity of the core genes in the group I strains of P . syringae; ß1 measures the difference in percent amino acid identity between the T3E families and the core gene families for the group I strains of P . syringae; ß2 , ß3 and ß4 measure the differences in percent amino acid identity for the core gene families for the legume pathovars of P . syringae , B . japonicum , and S . fredii , respectively , against that of the group I strains; ß5 , ß6 and ß7 allow the variation of the differences in percent amino acid identity between T3Es and core gene families across the groups; in particular , ß1+ß5 , ß1+ß6 and ß1+ß7 measure the differences in percent amino acid identity between the T3E families and the core gene families for the legume pathovars , B . japonicum , and S . fredii , respectively . An F test was used to test the null hypotheses that the percent amino acid identity for within-family comparisons between translated T3E and core gene sequences are equal within bacterial groups: ß1 = 0 , ß1+ß5 = 0 , ß1+ß6 = 0 and ß1+ß7 = 0 ( F test with degrees of freedom 1 and 83712 ) . A Bonferroni correction was used when applicable [70] . We used sequences of 30 confirmed functional tts-boxes from B . japonicum , S . fredii and M . loti MAFF303099 to train a Hidden Markov Model [13] , [30] , [71] . To identify candidate T3E genes , we identified CDSs downstream of tts-boxes with bit scores ≥5 . 0 , calibrated based on the identification of 11 functionally validated tts-boxes located on the pNGR234a plasmid [13] . To be considered , CDSs had to be encoded on the same strand as the tts-box , either up to 10 kb downstream or until another CDS on the opposite strand was encountered . TtsI-regulated operons , such as the nopB-rhcU operon of S . fredii NGR234 , can be substantial in length [72] . We used BLASTX ( e-value≤1×10−15 ) to filter out CDSs with translated sequences homologous to components of the T3SS , proteins encoded by organisms that lack a T3SS , or proteins with general housekeeping functions . We used BLASTN and sequences of candidate T3E-encoding genes to identify homologs from each of the eight genome sequences ( e-value cutoff≤1×10−15 ) . T3Es were grouped into families based on BLASTP scores ≤1×10−5 across ≥60% the length of the protein [38] . When all members of a family had amino acid identity ≥90% as determined using ClustalW , a single representative family member was chosen for testing [73] . In families of <90% amino acid identity , members representative of the diversity were tested . PCR , Gateway cloning into pDONR207 and the destination vector pDD62-Δ79AvrRpt2 , transformation into E . coli DH5α cells , and triparental mating into PtoDC3000 or ΔhrcC were done as previously described or according to the instructions of the manufacturer ( Invitrogen , Carlsbad , CA; [31] ) . Infiltration and HR assays were done as previously described [31] . Plants were grown in a controlled growth chamber environment ( 15-hour day at 22°C followed by 9-hour night at 20°C ) . Experiments were replicated a minimum of three times . | Rhizobia are an important group of bacteria that can enter into mutually beneficial symbiotic interactions with legume plants to fix atmospheric nitrogen . However , in order to do so , a complex dialog involving the exchange of chemical and molecular signals must occur between partners . Some species of beneficial rhizobia employ a type III secretion system , a well-characterized virulence mechanism used by pathogens to inject bacterial-encoded type III effector proteins directly into host cells to coerce the host into accommodating the microbe . In this study , we generated draft genome sequences and employed computational as well as experimental methods to identify type III effectors from eight strains representing Sinorhizobium fredii and Bradyrhizobium japonicum . We demonstrate that the type III effector genes of these rhizobial species are highly conserved in content with little diversity between strains . This work is an important step towards understanding the roles for type III secretion systems and their effectors in mutualistic interactions . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"gram",
"negative",
"plant",
"microbiology",
"microbial",
"evolution",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens"
] | 2013 | Mutualistic Co-evolution of Type III Effector Genes in Sinorhizobium fredii and Bradyrhizobium japonicum |
Because topical therapy is easy and usually painless , it is an attractive first-line option for the treatment of localized cutaneous leishmaniasis ( LCL ) . Promising ointments are in the final stages of development . One main objective was to help optimize the treatment modalities of human LCL with WR279396 , a topical formulation of aminoglycosides that was recently proven to be efficient and safe for use in humans . C57BL/6 mice were inoculated in the ear with luciferase transgenic L . major and then treated with WR279396 . The treatment period spanned lesion onset , and the evolution of clinical signs and bioluminescent parasite loads could be followed for several months without killing the mice . As judged by clinical healing and a 1 . 5-3 log parasite load decrease in less than 2 weeks , the 94% efficacy of 10 daily applications of WR279396 in mice was very similar to what had been previously observed in clinical trials . When WR279396 was applied with an occlusive dressing , parasitological and clinical efficacy was significantly increased and no rebound of parasite load was observed . In addition , 5 applications under occlusion were more efficient when done every other day for 10 days than daily for 5 days , showing that length of therapy is a more important determinant of treatment efficacy than the total dose topically applied . Occlusion has a significant adjuvant effect on aminoglycoside ointment therapy of experimental cutaneaous leishmaniasis ( CL ) , a concept that might apply to other antileishmanial or antimicrobial ointments . Generated in a laboratory mouse-based model that closely mimics the course of LCL in humans , our results support a schedule based on discontinuous applications for a few weeks rather than several daily applications for a few days .
Of the 350 million people exposed to the risk of Leishmania parasite inoculation and further development , 2 million each year experience the discomfort and potential complications of cutaneous leishmaniasis ( CL ) . Many active lesions are disfiguring , and remain so when healing as inesthetic scars that expose patients to social stigma , sometimes for life [1] , [2] . The demand for improved CL therapy has been fueled for decades by the lack of an efficient , affordable , easy-to-apply drug/schedule , as well as by the risks associated with the use of parenteral antiparasitic drugs such as pentavalent antimonial drugs or pentamidine [3] , [4] . Topical therapy of CL is a promising approach [5] , [6] . The aminoglycoside paromomycin is the most well studied compound as a potential topical treatment for CL [7] . First and second generation paromomycin-based ointments were either reasonably efficient [8] , [9] but too irritant ( first generation paromomycin-Methyl benzo chloride , “Leshcutan” ) [10] , [11] or well-tolerated but not efficient enough when first tested in humans ( second generation paromomycin-urea “WHO formulation” ) [12] , [13] . WR279396 , a third-generation aminoglycoside ointment that contains 15% paromomycin formulated in a hydrophilic vehicle as well as a second aminoglycoside , 0 . 5% gentamicin , was designed to be effective but non-irritative . This new formulation was recently shown to be efficient and safe for the treatment of L . major localized cutaneous leishmaniasis ( LCL ) ( Ben Salah , Buffet et al . , submitted and [14] ) . Although very encouraging , this result is only one step toward a simple and easily applicable therapy for this neglected disease . Various parameters such as frequency and duration of application or application in the presence or absence of an occlusive dressing-may markedly influence the efficacy or safety of topically applied formulations [12] , [13] , [15] , [16] . For example , once-a-day applications of Leshcutan are associated with less frequent and less severe local reactions than a twice-a-day application schedule [17] . Though still suboptimal , a 28-day schedule of paromomycin-urea ( WHO formulation ) is significantly more efficient than a 14-day schedule [12] . These 2 examples show that optimizing application parameters through clinical trials , the most reliable approach , takes years . Also , for obvious ethical reasons , there is usually no untreated control group in clinical trials , making interpretation of the mechanisms of drug action more difficult . In order to more rapidly and accurately identify important parameters that influence the efficacy of WR279396 , we designed and used a mouse model of CL that mimics important features of the natural sand fly dependent-transmission of parasites to mammal . A relatively low ( 104 ) inoculum of L . major metacyclic promastigotes was injected in the C57BL/6 ear dermis [18] . As in a majority of patients with L . major CL [19] , the development of localized dermal lesions in C57Bl6 mice is followed by spontaneous healing over the course of weeks to months [18] . Because luciferase transgenic parasites were used in this model , the kinetics of parasite load could be established without killing the mice: indeed , a linear correlation between bioluminescence values and parasite loads assessed by the reference limiting dilution technique has been previously established [20] .
Female C57BL/6 ( 5 week old ) and Swiss nu/nu mice were purchased from Charles River ( Saint Germain-sur-l'Arbresle , France ) and were housed under institutional guidelines of the A3 Animal facility at Institut Pasteur ( Paris , France ) . A 1 . 66 kbp firefly luciferase coding region was cut from pGL3 basic ( Promega , Madison WI ) using NcoI/EagI and subsequently cloned into the Leishmania expression vector pF4X1 . HYG ( Jenabioscience , Jena , Germany ) with a marker gene for selection with Hygromycin B ( Cayla , Toulouse , France ) which was previously cut with NcoI/NotI . In this construct , 18s rRNA flanked the luciferase and HYG genes . Following linearization with SwaI , luciferase and HYG genes were integrated into the 18s rRNA locus of the nuclear DNA of Leishmania . Transfections were realized by electroporation with the following conditions: 25 µF , 1500 V , in 4 mm cuvette; 3 . 75 kV/cm [21] . Following electroporation , cells were incubated 24 h in media without drug and plated on semisolid media containing 100 µg/ml of hygromycin B [20] . Transgenic luciferase L . major strain NIH173 ( MHOM/IR/-/173 ) amastigotes were isolated from infected Swiss nude mice . Briefly , the promastigote developmental stage was grown at 26°C in M199 media supplemented with 10% FBS , 25 mM Hepes pH 6 . 9 , 12 mM NaHCO3 , 1 mM glutamine , 1×RPMI 1640 vitamin mix , 10 µM folic acid , 100 µM adenosine , 7 . 6 mM hemin , 50 U/ml of penicillin and 50 µg/ml of streptomycin [21] . Infective-stage metacyclic promastigotes were isolated from stationary phase cultures ( 6 day old ) using density gradient centrifugation , as previously described [22] . C57BL/6 mice were anaesthetised by intraperitoneal administration of a mixture of Ketamine ( 120 mg/kg−1 Imalgene 1000 , Merial , France ) and Xylazine ( 4 mg kg−1; Rompun 2% , Bayer , Leverkusen , Germany ) . Ten thousand metacyclic promastigotes per 10 µl of Dulbecco's phosphate buffered saline ( PBS ) were injected in the right ear dermis . Images of ketamine-xylasine anaesthetised mice were captured each day bioluminescence analyses were performed . The clinical features of parasite-loaded ear were examined based upon three phases: 1 ) early , leucocyte infiltrate-free inflammatory , processes , 2 ) leucocyte infiltrates-positive inflammatory processes and 3 ) late repair processes could be distinguished . Only one name , “lesion” , was used to designate these different processes . The “lesion” size measurement ( mm2 ) was approximated from the picture by fit within a rectangle . Luciferin ( D-Luciferin potassium salt , Xenogen , California ) , the luciferase substrate , was intra-peritonealy inoculated into mice at a concentration of 150 mg/kg 25 minutes before bioluminescence analysis . Mice were anaesthetised in a 2 . 5% isoflurane atmosphere ( Aerane , Baxter SA , Maurepas , France ) for 5 minutes and maintained in the imaging chamber for analysis . Emitted photons were collected by 1 minute acquisition with a charge couple device ( CCD ) camera ( IVIS Imaging System 100 Series ) using the high resolution ( small bining ) mode . Analysis was performed after defining a region of interest ( ROI ) that delimited the surface of the entire ear . The same ROI was applied to every animal at every time point . Total photon emission from the ventral image of each mouse ear was quantified with Living Image software ( Xenogen Corporation , Almeda , California ) , and results are expressed in number of photons/sec/ROI . The photon signal from the ear is presented as a pseudocolor image representing light intensity ( red = most intense and blue = least intense ) and superimposed on the grey scale reference image . Of note , the lower threshold bioluminescence value indicates a parasite load of ≥5000 parasites per ear , precluding any detection of persisting parasite population that oscillates between 500 and 1000 parasites . Forty to 60 animals per experiment were inoculated with transfected L . major and the total photon emission of each ear was quantified 11 days later . Mice were monitored and distributed in groups according to an equal median bioluminescence value ( 1×106–5×106 photons/sec/ROI ) and standard deviation . Each experimental group contained 7 to 10 mice , each individually ear-tagged ( the contralateral ear with respect to the inoculation site ) . Topical formulations were prepared at the Walter Reed Army Institute of Research ( Washington DC ) . WR279396 consists of paromomycin sulphate ( 15% ) plus gentamicin ( 0 . 5% ) in a vehicle as previously described [23] . From day eleven post-L . major inoculation , topical ointments were applied to parasite-loaded ears once every two days for 10 days or once everyday for 5 days . Each formulation was applied using a sterile tip directly onto the ears to form a thin layer . Control groups were treated with the vehicle used in the medication without any of the active ingredients , i . e . , the paromomycin and gentamicin . The ointment was either left open without dressing or covered with an occlusive dressing . The occlusive dressing was an adhesive polyurethane membrane ( Tegaderm; 3M Health Care , St Paul , USA ) that keeps water but is permeable to both water vapour and oxygen . Then two independent leaflets of 3M Micropore Surgical Tape ( 3M Health Care ) were directly applied to the Tegaderm . This tape permitted maintenance of Tegaderm and formulation in contact with the ear during the two days . We estimated the number of parasites present in parasite-loaded ears as previously described [24] . Ears were cut off . The dorsal ear half was separated from the cartilage-containing ventral ear half , cut into small pieces and ground in HOSMEM-II culture medium using a glass tissue homogenizer . Tissue/organ homogenates were serially diluted in HOSMEM-II culture medium and then dispensed into 96-well plates containing semi-solid agar ( Bacto-Agar , Difco , Detroit , MI ) supplemented with 10% sterile rabbit blood collected on heparin . Plates were incubated for ten days and each well was then examined and classified as positive or negative according to whether or not viable promastigotes were present . Limiting dilution analysis was then applied to the data to estimate the number of viable parasites , expressed in limiting dilution assay units ( LDAU ) [25] . Statistical analysis of the results was based on the maximal likelihood method [26] , [27] . Lesion size or log transformed parasite loads were analyzed with a two-way analysis of variance ( ANOVA ) . The two factors examined were the treatment ( untreated , vehicle , drug , drug with occlusion ) and the period of observation ( treatment , post-treatment and final ) in the statistical environment R . The assumption of homoscedasticity and normality were tested with the Bartlett and Kolmogorov-Smirnov test , respectively . If the interaction term was significant , pair wise comparisons using t tests were realized for each combination of factors . A probability level of p<0 . 05 was accepted for the purpose of declaring statistically significant treatment effects .
The first objective of this study was to design and validate standardized readout assays for assessing different drug regimens using C57BL/6 mice inoculated with Leishmania major . To carry out these experiments , 104 luciferase-expressing L . major metacyclic promastigotes were inoculated intradermally into the mouse ear . Parasites produced a significant bioluminescent signal in situ allowing parasite load expansion and reduction to be monitored non-invasively . The development and outcome of parasite burden and parasite-loaded ear features were examined simultaneously over a period of 3 months . The relationship between bioluminescence and the clinical features of the ear were respectively assessed by quantifying the number of photons per second per ear and measuring the “lesion” area . Figures 1A and 1B illustrate the real-time bioluminescent images and clinical signs displayed by the L . major-inoculated ear from a representative C57BL/6 mouse ( untreated group ) . The first post-inoculation phase ( days 0–11 ) was characterized by a sharp increase of the bioluminescent signal at the inoculation site ( from 7×103 to 4 . 4×106 photons/sec/ear at day 11; Figure 1B , C ) . By day 7 , mouse ears displayed no clinically detectable sign ( Figure 1A ) . However , a leukocyte infiltrate-free tiny red spot ( 5 mm2 ) was observed at day 11 ( Figure 1A , C ) . Thus , during the first stage of parasite development no significant correlation was found between the bioluminescence value at the inoculation site and the clinically detectable features . By day 22 , the parasite load peaked ( Figure 1B and 1C ) with a median value of 1 . 5×107 photons/sec/ear which was associated with the first bona fide cutaneous clinical signs ( Figure 1A and 1C ) . The next phase of L . major-driven processes was characterized by a relatively sharp decrease in bioluminescence followed by healing of the ear lesion ( Figure 1A , 1B and 1C ) . Following the complete and stable healing of this dermal lesion , no more bioluminescent signal was detected in the ear tissue ( Figure 1B and 1C ) . We are aware that any persisting parasite load with a population size value ≤5000 per ear is not detectable using bioluminescence: thus , between days 80–96 post inoculation at the time of mouse sacrifice , mouse ears were recovered in the control and treated group . Using the LDA readout assay , these ears were monitored for the presence of persisting parasites . 40% of the ears were positive in all groups ( ≤500 parasites per ear ) and these percentages were obtained from two independent experiments ( data not shown ) . In contrast , the persistent presence of a low number of parasites as measured by LDAU was noted in the inoculation site ( 3 positive mice out of 7 ) for up to 80 days post-inoculation . These measurements helped us to define the onset of the first topical ointment application ( WR279396 vehicle or WR279396 ointment with or without dressing ) . We decided to initiate treatment at day 11 post-inoculation for three reasons . First , at this time point , a high parasite load ( bioluminescence values>1×106 photons/sec/ear ) was reproducibly measured . Secondly , these values were observed in the median part of acute-phase load , allowing for monitoring of either an increase in parasite load in the absence of any topical application or a decrease in treated groups . Thirdly , the last topical ointment application in the group of mice treated with WR279396 was coincident with the highest parasite load measured in the control group . By day 11 post-parasite inoculation , C57BL/6 mice were distributed in different groups on the basis of equal median bioluminescence values . WR279396 was applied topically to the L . major-inoculated ear every two days for 10 days . Occlusive dressing was performed by covering the L . major-loaded ears with adhesive polyurethane dressing ( Tegaderm ) and a surgical tape to maintain the formulation for 2 days ( Figure 2 ) . An evaluation of the effect of WR279396 with an occlusive dressing was monitored by measuring the bioluminescence and ear “lesion” area ( Figures 3B and 3C ) . Three periods of observation have been defined i ) the 10-day treatment period ii ) the post-treatment period , which ends with the absence of any bioluminescence signals in the control group and iii ) the late period . As controls , three groups were analysed . In the first group , ears were left untreated . In the second group , the paromomycin- and gentamicin-free vehicle was applied to L . major-inoculated ears that remained uncovered after application . In the third group , the WR279396 vehicle was applied and ears were immediately covered with an occlusive dressing . In all control groups , parasite load as well as lesion onset development and healing were simultaneously assessed . No statistical differences in parasite loads and lesion area were observed in any period between untreated and ointment vehicle-treated groups regardless of the period under study with respect to the measurement ( not shown ) . Monitoring of bioluminescence values showed that topical treatment with WR279396 ( without a dressing ) accelerated the decrease of both the parasite load ( Figure 3A ) and “lesion” area ( Figure 3B and 3C ) . Two-way ANOVA analysis indicated a significant effect of treatment ( P-value<9 . 2×10−6 ) and period ( p-value<2 . 2×10−6 ) on parasite load for the whole experimental group , and there was a significant interaction between treatment and period effect ( p-value<0 . 008 ) . The parasite load ( grey line; Figure 3A ) decreased rapidly after the fourth application ( day 18; Figure 3A ) . Median values of bioluminescence indicated that parasite loads in the group of mice left without dressing were significantly lower than the control group during the treatment period ( Figure 3A and 3D; grey line vs blue line and box plot-: p-value = 0 . 0033 ) . Of note , during the post-treatment period , a rebound pattern of parasite load was observed in mice treated with ointment without occlusive dressing ( grey line; Figure 3A ) no statistical difference between groups being noted ( Figure 3D ) . For mouse ears that were covered with WR279396 under an occlusive dressing , mean parasite loads ( Figure 3A; brown lines ) decreased earlier than those with WR279396 left without any dressing . One day post the last application , bioluminescence values reached threshold bioluminescence values ( Figure 3A ) in 80% of mice ( 8 out of 10 ) . During the post-treatment period , statistical analyses indicated that parasite load in the group of mice with an occlusive dressing was significantly lower than in the group of mice treated without a dressing ( Figure 3B , p-value = 2 . 2×10−8 ) . The higher significant therapeutic effect of the drug in the presence of an occlusive dressing during this post-treatment period is illustrated in figures 3B , 3C ( day 28: p-value = 0 . 00055 ) . Furthermore , no rebound of parasite load was observed in this group . In conclusion , the decrease in parasite loads and the healing process occurred earlier in mice treated with WR279396 under an occlusive dressing . Among this group of mice , neither clinical relapse-as measured by leucocyte infiltrate-related “lesion” area-nor rebound of parasite load was detected . Parasite rebound was observed in some mice given WR279396 without occlusive dressing . The individual follow-up of parasite-loaded mouse ears in real time indicated a clear dichotomy in the patterns of parasite load outcome ( Figure 4; same experiment as Figure 3; n = 10 ) . In the majority of treated mice , parasite load decreased faster than in control mice with a bioluminescence value lower than 1×106 photons/sec/ear ( Figure 4A; 6 out 10 mice depicted in green ) at day 33 . Furthermore , no rebound was detected in this group of mice ( Figure 4B ) . In contrast , bioluminescence values for the four remaining mice were higher than 1×106 photons/sec/ear at day 33 ( Figure 4A and 4B; 4 mice depicted in red ) : post treatment , either the parasite load reduction pattern followed the same profile ( 1/4 mice; Figure 4B ) as the one displayed by mice of the control group ( grey area ) or relapsed occurred at day 22 ( 3/4 mice; Figure 4B ) . The parasite load in this latter group of 3 mice remained higher than the parasite load of the control group from days 30 to 60 . Interestingly , bona fide lesion area values did not assess any obvious clinical failure except in 1 mouse which harboured the highest parasite load ( see arrow in Figure 4B ) and displayed a somewhat slower healing process ( Figure 4C; see arrow in Figure 4C ) . These data suggest that i ) the rebound pattern of parasite load , which was observed in mice treated in the absence of occlusive dressing , had a clinical impact in a minority of mice and ii ) the bioluminescence imaging data provided relevant information on parasite load fluctuations that were not provided by careful clinical monitoring . We also monitored application regimens of WR279396 to determine which one might have a superior therapeutic index against the parasite . The experimental protocol shown in figure 5 was as previously described except for a different schedule of drug ointment application . At day 11 , the topical ointment was applied on parasite-loaded ears either daily for 5 days or once every two days for 10 days . A control group ( no medication ) was used in parallel for determining comparability and efficacy of the different topical therapy regimens . As previously described , all parasite-loaded ears exposed to five applications for 10 days were healed by day 21 ( Figure 5A ) without relapse by day 64 . All lesions ( 7/7 ) treated with WR279396 daily for 5 days had healed at day 21 ( end of the topical therapy ) , but 71% ( 5/7 ) and 14% ( 1/7 ) of mice relapsed at day 50 and day 60 , respectively . The clinical aspect of the lesions at day 36 post-inoculation ( Figure 5B ) confirmed the greater efficiency of the 5 every two days application over a 10 day schedule . By two-way ANOVA , it was shown that the difference between treatments depends on the observation periods considered ( significant interaction term , p-value<0 . 05 ; Figure 5C ) . Pair wise comparisons using t-tests for each combination of factors are shown in figure 5D . The integrative analysis of parasite load evolution in 4 experiments , involving mice receiving 5 applications for 10 days either with ( n = 31 ) or without ( n = 23 ) an occlusive dressing , shows that 74% of mice treated without a dressing controlled parasite loads ( 17/23 ) without relapse . Of the 6 remaining mice , 2 were unresponsive , as shown by parasite load values similar to untreated mice . The other mice initially controlled parasite loads and lesion size during the treatment period , but relapsed by day 30 as shown by parasite load values similar to or higher those measured from the ears of untreated mice . In contrast , 94% of mice treated with an occlusive dressing healed ( 29/31 ) by day 30 . Only 6% ( 2/31 ) had detectable-though very low-parasite load ( bioluminescence level<1×106 photons/sec/ear ) during follow-up . These results allow us to establish the greater parasitological efficacy of the schedule using an occlusive dressing , with a trend toward a prophylactic effect on relapse after a successful course of WR279396 .
WR279396 , a third-generation aminoglycoside-based ointment , was efficient on L . major-induced localized cutaneous lesions ( LCL ) in C57BL/6 mice . Five applications for 10 days under occlusion induced a 94% healing rate by day 30 , without re-expansion of parasite loads . This high cure rate , as well as the general evolution profile in both treated and control mice , is strongly reminiscent of what has been observed in clinical trials ( Ben Salah , Buffet et al . -submitted and [14] ) , providing a strong validation of this new model for drug-testing purposes . The adjuvant use of an occlusive dressing significantly enhanced control of parasite loads . Several non-mutually exclusive mechanisms may account for these effects . First , the dressing prevented removal of the ointment from the lesion by protecting the skin from scratching , rubbing and scraping . These latter observations have been made in patients treated without occlusion , an important proportion of the ointment being wiped off by clothes during the day , sheets during the night or even attracted to a “protective” gauze put on the top of ulcerated lesions . Second , occlusion on burns or wounds favours epidermal regeneration ( ie , ulceration closure ) . Finally , water retention by semi-permeable occlusive dressing ( like the polyurethane film used here ) results in hydration of the ointment application zone [28] and likely improves the penetration and diffusion of hydrophilic antiparasitic compounds into the dermis [15] , where intracellular amastigotes multiply . The aminoglycosides paromomycin and gentamicin , the active ingredients in the WR279396 ointment , are OH-rich hydrophilic compounds . Whether the dominant mechanism of the adjuvant effect is merely mechanical ( enough ointment maintained on the lesion ) or linked to dermal diffusion issues , the occlusive dressing enhanced the healing process induced by active ingredients , and prevented persisting parasite loads to re-expand . Now that this adjuvant effect is established , future studies should be set up for dissecting its fine mechanisms . Apart from a mild difference in thickness , LCL lesions in mice resemble human lesions both clinically and histologically . As opposed to systemic drug testing , topical drug testing in mice will be relevant since potential pharmacokinetic differences between mouse and human skin are expected to be minor and easily tractable to further analyses . These observations , along with the careful validation of the model , support the assumption that our results are likely to apply to human therapy . To our knowledge , occlusion has never been fully validated as an adjuvant for the topical therapy of human cutaneous lesions driven by invasive microorganisms , but several case reports have proposed this approach both in CL [29] , [30] and non-infectious dermatologic conditions [15] . We provide here a strong validation of a concept that might apply to other antileishmanial or antimicrobial ointments . Ointments are usually painless , their application requires no sophisticated expensive device or local anaesthesia and they can be applied easily to both children and adults by a primary care health provider with minimal training . Excluding those L . braziliensis . foci where the incidence of mucosal extension can be high , ointment therapy of cutaneous lesions otherwise declared as neglected diseases should be favoured since the potential adjuvant effect of occlusion might help some ointment formulations to reach the required efficacy for development . The duration of treatment is another important determinant of reaching a stable cure . A very short 5-day daily application schedule under occlusion led to a “rebound” pattern similar to that displayed in mice treated for 10 days without occlusion . In other words , too short of an application period may lead to parasite load rebound , this latter risk being partially controlled by an occlusive dressing . The ability to perform individual mouse follow-up revealed a dichotomic pattern of parasite load evolution ( “sustained control” versus “unstable control with parasite rebound” ) , pervasive over many weeks post transient topical ointment application . Interestingly , these patterns were displayed over several weeks , i . e . , well beyond the treatment application period . Parasite load level at the end of applications was not a good predictor of further evolution ( Figure 4 ) . So , not only parasite killing but also some modification of parasite environment determined the long-term outcome of tissue damage and repair processes . It is then very likely that , during the treatment application period , some integrated programs are triggered that will be the dominant determinant of evolution . Those results fit well with observations in human CL , such as the low prognosis value of parasitological tests at the end of therapy or the efficacy of therapeutic schedules stopped before lesion healing [4] , [31] . Taken together , these experiments show that parasites must be exposed to the drug for>5 days to drive evolution toward long term sustained control of parasite loads and clinical healing . Duration of drug exposure was a stronger determinant of outcome than the total amount of drug used . Intracellular pharmacokinetics of aminoglycosides helps understand the mechanism leading to this observation . In eukaryotic cells exposed to aminoglycosides in vitro , a slow ( 2–4 days ) lysosomal accumulation is observed , followed , when aminoglycosides are removed from the extracellular medium , by an even slower ( 2–5 days ) release [32] . Interestingly , the lysosome is the only subcellular compartment in which aminoglycosides accumulate , an important feature of their antileishmanial efficacy [32] , [33] . So , provided that appropriate concentrations of aminoglycosides are reached in the dermal intercellular space , relatively discontinuous applications would probably suffice to allow intracellular killing of replicating amastigotes and long term sustained control of parasite loads . Taken together , our observations will help select the most efficient ointment application schedules for implementation , in the context of the therapy of this neglected disease , by health care providers with little resources and heavy duties . Even relatively discontinuous applications for a few weeks should be preferred to many daily applications for a few days . Our model offers relevant preclinical readout assays i ) of the efficacy of a topical ointment delivered under occlusion or not ii ) for establishing the proper regimen/schedule that allows sustained parasite load reduction and lesion healing during post-treatment period features . This luciferase-based imaging study might be useful for pre-clinical evaluation of novel formulations containing molecules that target parasite-loaded cells residing in the dermis as well as molecules that contribute to damaged skin-repair processes . Next challenges will be to screen molecules expected to act on the amastigote population that persist in the dermis or in distant sites [34] and to investigate the acquisition in real time of long-term protective immunity . | When initiating the cutaneous disease named cutaneous leishmaniasis ( CL ) , Leishmania parasites develop within the parasitophorous vacuoles of phagocytes residing in and/or recruited to the dermis , a process leading to more or less chronic dermis and epidermis-damaging inflammatory processes . Topical treatment of CL could be a mainstay in its management . Any improvements of topicals , such as new vehicles and shorter optimal contact regimes , could facilitate their use as an ambulatory treatment . Recently , WR279396 , a third-generation aminoglycoside ointment , was designed with the aim to provide stability and optimal bioavailability for the molecules expected to target intracellular Leishmania . Two endpoints were expected to be reached: i ) accelerated clearance of the maximal number of parasites , and ii ) accelerated and stable repair processes without scars . A mouse model of CL was designed: it relies on the intradermal inoculation of luciferase-expressing Leishmania , allowing for in vivo bioluminescence imaging of the parasite load fluctuation , which can then be quantified simultaneously with the onset and resolution of clinical signs . These quantitative readout assays , deployed in real time , provide robust methods to rapidly assess efficacy of drugs/compounds i ) to screen treatment modalities and ii ) allow standardized comparison of different therapeutic agents . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/antimicrobials",
"and",
"drug",
"resistance"
] | 2007 | Optimization of Topical Therapy for Leishmania major Localized Cutaneous Leishmaniasis Using a Reliable C57BL/6 Model |
The spinal cord participates in the execution of skilled movements by translating high-level cerebral motor representations into musculotopic commands . Yet , the extent to which motor skill acquisition relies on intrinsic spinal cord processes remains unknown . To date , attempts to address this question were limited by difficulties in separating spinal local effects from supraspinal influences through traditional electrophysiological and neuroimaging methods . Here , for the first time , we provide evidence for local learning-induced plasticity in intact human spinal cord through simultaneous functional magnetic resonance imaging of the brain and spinal cord during motor sequence learning . Specifically , we show learning-related modulation of activity in the C6–C8 spinal region , which is independent from that of related supraspinal sensorimotor structures . Moreover , a brain–spinal cord functional connectivity analysis demonstrates that the initial linear relationship between the spinal cord and sensorimotor cortex gradually fades away over the course of motor sequence learning , while the connectivity between spinal activity and cerebellum gains strength . These data suggest that the spinal cord not only constitutes an active functional component of the human motor learning network but also contributes distinctively from the brain to the learning process . The present findings open new avenues for rehabilitation of patients with spinal cord injuries , as they demonstrate that this part of the central nervous system is much more plastic than assumed before . Yet , the neurophysiological mechanisms underlying this intrinsic functional plasticity in the spinal cord warrant further investigations .
Results from a plethora of studies clearly indicate that the learning of new motor skills in humans induces functional plasticity in a distributed network of brain areas [1–4] . Despite such advances in our knowledge base , the current understanding of the neural substrates mediating motor skill learning remains limited , as contribution of the spinal cord to this memory process is still unaccounted for . Previous animal research has documented the existence of long-term plasticity in the spinal cord following basic and primitive forms of learning [5 , 6] . For instance , using both lesion [7] and operant conditioning [8 , 9] paradigms , such studies have demonstrated that motor memories can be stored within the spinal circuits after extended practice . Similarly in humans , long-term practice of skilled movements has been shown to diminish the amplitude of spinal reflexes in muscles of highly trained individuals ( e . g . , ballet dancers [10] and volleyball players [11] ) as compared to control subjects . Yet , it is unclear whether the spinal cord manifests plastic changes during early stages of motor learning [12 , 13] and whether it plays an active role in the initial acquisition of new motor skills . Consequently , human motor learning models usually consider the spinal cord as a passive relay of information from the brain ( controller ) to the muscles ( effectors ) , with no active learning-related role entrusted to the spinal circuitry [12] . This view mainly stems from a computational perspective of motor learning , which assumes that circuits at the spinal cord level , on which cortical plasticity is grounded and stabilized , are hardwired [6] . Contrary to this notion , new lines of evidence at the cervical [14] and lumbar [15 , 16] spinal levels demonstrate that spinal reflex activities can be selectively modulated by short periods of motor learning , hence suggesting that the human spinal cord may be actively involved during the learning of novel motor skills . For instance , in an innovative electrophysiological study , Meunier et al . [15] showed that homosynaptic depression in soleus muscle ( i . e . , a measure of local depletion in primary afferent neurotransmitters ) significantly changed following adaptation to a complex pattern of changes in resistance during stationary cycling . This finding suggests that the pattern of sensory inflow plays an essential role in producing plasticity at the level of spinal cord . Despite such evidence , an important limitation of many studies using electrophysiological recordings has been that they do not allow one to distinguish plasticity in the spinal cord from functional change caused by descending cerebral inputs . Thus , whether intrinsic plasticity occurs in the human spinal cord during acquisition of new motor skills remains an open question . One way to overcome this limitation is to record both the brain and spinal cord activities simultaneously in order to assess the extent to which changes in the spinal cord activity correlate with or are statistically independent from the plastic changes that happen at the brain level during motor learning , hence allowing the identification of intrinsic changes at the spinal cord level . For the first time , we used functional magnetic resonance imaging ( fMRI ) to test the hypothesis that human spinal cord activity at the cervical level shows intrinsic learning-related changes during motor sequence learning ( MSL ) . To do so , we acquired simultaneous images of the entire brain and cervical spinal cord during performance of a motor sequence task in healthy young subjects ( Fig 1A–1C ) . A well-known MSL paradigm was chosen so that we could then link the novel imaging findings found here at the spinal cord level to the well-established behavioral determinants and neural correlates of MSL at the brain level ( e . g . , see [2 , 17] , for reviews ) . This allowed us to examine the relative contribution of cortical , subcortical , and spinal regions in modulating performance during the early acquisition phase of a new sequence of movements and to investigate the functional connectivity between these structures over the course of motor learning .
We used a specific slice prescription approach to acquire functional scans of the whole human brain and cervical cord during performance of an MSL task ( Fig 1A–1C ) . We trained participants ( n = 25 , median age = 24 . 0 y; number of males: 11 ) to carry out a given self-generated five-element finger sequence ( Fig 1A , complex sequence [CS] ) and a matched control motor sequence task ( Fig 1A , simple sequence [SS] ) with their nondominant left hand . They were required to perform the finger movements in the CS and SS conditions as quickly and accurately as possible and to make as few errors as possible . The CS and SS conditions were split evenly across blocks in a pseudorandom fashion and were flanked by blocks of rest periods ( Fig 1B ) . The overall mean error rate was 3 . 58% , with 3 . 2% errors committed out of all trials during the CS and 3 . 95% errors during the SS condition . The Wilcoxon Signed Ranks tests indicated that there was no statistically significant difference in error rates between the two conditions ( p = 0 . 56 ) . An ANOVA for repeated measures on performance speed across all blocks in both CS and SS conditions revealed that learning curves differed across the CS and SS conditions ( Fig 1D; significant condition × blocks interaction; F11 , 264 = 1 . 92 , p<0 . 05 ) . This suggests that distinct learning mechanisms were involved during the SS and CS training conditions , the former being due to a general improvement in motor performance by repeated practice and the latter resulting from sequence-related improvements in motor performance in addition to the nonspecific motor practice effect [18] . Although , subjects’ finger movements became faster by the end of training in both conditions ( mean duration of the last two blocks compared to the first two blocks in the CS , t24 = 11 . 0 , p<0 . 001 , and in the SS , t24 = 7 . 7 , p<0 . 001 ) , there was significantly greater improvement in speed over the course of learning during the CS condition compared to the SS condition ( Fig 1E , t24 = 2 . 11 , p<0 . 05 ) . However , the average performance speed was significantly higher in the SS condition compared to the CS condition ( main effect of condition; F1 , 24 = 9 . 4 , p<0 . 01 ) . We first assessed fMRI correlates of motor practice during the SS ( blue ) and CS ( red ) conditions compared to the baseline rest periods in the brain and cervical cord separately using two repeated measures general linear models ( GLM ) at the group level . All the brain and spinal cord group-level activation maps were corrected for multiple comparisons using Gaussian random field ( GRF ) theory correction . As expected at the brain level , similar sensorimotor regions as previously reported ( cerebellum , putamen , supplementary motor area , premotor and primary sensorimotor cortices ) were activated during the CS and SS conditions ( S1 Fig ) [1 , 2 , 4] . Most importantly , however , greater blood oxygenated level–dependent ( BOLD ) activity was also found at the expected C6–C8 spinal levels of the cervical cord , wherein motoneurons innervating the finger muscles reside , during performance of the CS and SS conditions ( Fig 2A; main effect of practice , corrected cluster-level p-values using GRF: p < 0 . 0001 for CS and p = 0 . 001 for SS condition ) . Also as expected , BOLD activity in this spinal cord region was mostly located on the side ipsilateral to the hand ( left ) used to perform the tasks . As shown in Fig 2A , group-level activated regions related to the main effect of practice during the CS and SS conditions were almost overlapping ( i . e . , peaks of activity in both conditions were located ipsilaterally at the C6–C8 spinal level ) . However , critically , individual subject’s activation maps related to the CS condition ( S2 Fig ) also revealed that 21 out of 23 subjects showed a consistent cluster of activity in that same region , hence demonstrating the robustness of this pattern of activity at the spinal cord level . Furthermore , since the activation maps were largely overlapping between the two conditions , a direct comparison between them did not result in any significant cluster at the group level using the applied family-wise corrected threshold . To examine further the differences between the effects of CS and SS conditions ( due to the acquisition of a new motor sequence and the complexity of the task itself ) on BOLD changes in the spinal cord , we employed a region-of-interest ( ROI ) -based analysis to measure changes in the spatial extent and amplitude of the activated voxels in each participant . For each subject and condition , the ROI was selected as a set of activated voxels ( Z > 2 . 3; p < 0 . 01 uncorrected ) within the C6–C8 spinal level ( see S3 Fig for the applied anatomical mask ) from the individual-specific activation map of that condition . The BOLD signal amplitude changes in the CS and SS conditions were 2 . 9 ± 0 . 6% and 2 . 4 ± 0 . 5% , respectively ( mean ± SEM; Fig 2B; with an average difference between the CS and SS conditions of 0 . 5 ± 0 . 18% ) . The average numbers of activated voxels within the C6–C8 cervical segments during the CS and SS conditions were also 55 ± 5 and 44 ± 4 . 5 , respectively ( mean ± SEM; Fig 2C ) . Thus , the BOLD signal amplitude and spatial extent were significantly larger in the CS condition compared to the SS condition ( repeated-measure t test; t22 = 2 . 62 , p<0 . 05 for amplitude and t22 = 2 . 57 , p<0 . 05 for spatial extent ) . It is important to note that the latter analysis ( i . e . , spatial extent ) was performed on preprocessed data that were not spatially smoothed in order to yield independent measures of activation amplitude and spatial extent ( i . e . , the larger number of activated voxels in CS was not a side effect of spread of activity into the surrounding voxels due to spatial smoothing ) . Furthermore , we investigated the center of gravity ( mean coordinates and standard deviations ) of the activated voxels in the CS and SS conditions along the different axes ( x , y , and z ) to identify the direction ( s ) of expansion of activity in the CS condition compared to the SS condition . Although the mean of the center of gravity across the subjects was similar in both the CS and SS conditions along all three axes ( p > 0 . 35; paired sample t test ) , the spatial extent of activated voxels along the dorsoventral axis of the spinal cord was larger in the CS condition than in the SS condition ( the standard deviation of the y-coordinate of activated voxels in each subject was on average 11% larger in the CS condition than in the SS condition; p = 0 . 055 , t = 2 . 03 , paired sample t test ) . Also , the standard deviation of activated voxels was , respectively , 7% and 2% larger in the rostrocaudal and mediolateral directions in the CS condition than in the SS condition , but it did not reach significance ( p = 0 . 32 and p = 0 . 48 , respectively ) . To test more specifically whether the observed cervical BOLD activity was modulated by the amount of motor learning in each subject , we sought to identify the neural correlates of improvement in performance speed using a repeated-measures general linear model . The results revealed that the amount of improvement in motor performance during the CS condition , but not the SS condition , was correlated with BOLD activity changes within two spinal cord clusters located at the same spinal level as those observed in the analysis looking at the main effect of practice ( Fig 3 , spinal level , corrected cluster-level p-values using GRF: p < 0 . 01 for both clusters ) . Here , the activity was primarily located bilaterally in the intermediate part of C7 and in the ipsilateral C8 region . Importantly , we also investigated the difference in modulation by performance speed across conditions ( speed performance by condition interaction ) . Similar to the results of the CS modulation alone reported above ( see Fig 3 ) , this analysis resulted again in two significant activation clusters centered at the C7 and C8 spinal levels ( S4 Fig ) , hence suggesting that , compared to the SS condition , the CS condition generated greater modulated activity during motor learning in these particular regions of the spinal cord . Finally , it is noteworthy that the performance speed was significantly increased in both CS and SS conditions over the course of learning ( p<0 . 001 in both condition; Fig 1D ) . Thus , this suggests that the lack of modulation in spinal cord activity during the SS condition could not be due to a thresholding effect . This also suggests that motor speed change alone was insufficient to account for the activation changes observed in the spinal cord over the course of learning . Finally , we also conducted a similar analysis at the brain level using standard preprocessing and statistical methods . As expected ( see [2 , 17] ) , we found clusters of activity in motor-related cortical regions including the hand-related area of the contralateral sensorimotor cortex ( M1/S1 areas ) and the dorsal premotor cortex , as well as two other clusters in subcortical regions including the contralateral putamen and the ipsilateral lobule V-VI of the cerebellum , all of which showed significant learning-related modulation in activity during the CS condition ( Fig 3 , cortical and subcortical levels , corrected for family-wise error using GRF , cluster significance threshold of p < 0 . 01; S2 Table ) . By contrast , and importantly , there was no brain area showing significant modulation in activity with performance speed during the SS condition . The classical fMRI analysis employed so far showed that changes in cervical cord activity were linked to learning-related behavioral improvement , particularly in the CS condition . Yet , this approach cannot inform as to whether these changes are a simple reflection of plasticity occurring within supraspinal sensorimotor structures or whether they actually represent intrinsic local plasticity at the spinal cord level . To further explore this issue , we conducted a conditional GLM analysis based on partial correlations [19 , 20] to account for and partial out the possible supraspinal contribution ( conditioning variables ) onto the cervical cord BOLD activity ( dependent variable ) when modeling the learning-related improvements in behavior ( independent variable ) . We tested two conditional models of spinal activity . In the first model , we accounted for activities of the main brain areas that are known to send direct and indirect efferent to the spinal cord , hence influencing spinal cord excitability [21–23] . These areas included the contralateral M1 , dorsal and ventral premotor cortices , supplementary motor area , anterior cingulate , S1 , and ipsilateral cerebellum ( S1 Table ) . In the second model , we incorporated , as conditioning variables , all brain areas that showed learning-related changes in activity ( Fig 3 , cortical and subcortical levels; S2 Table ) . In each model , we removed the effects of activity within the conditioning brain areas on the cervical activity and investigated the remaining learning-related modulation within the spinal cord . Interestingly , the results ( see Fig 4A and 4B ) revealed significant changes in cervical activities located at the C7–C8 segments that were modulated in association with behavioral improvements and were independent of concomitant signals from brain structures known to project to the spinal cord ( first model ) or from brain areas that showed learning-related activity changes ( second model ) , respectively . The similarity between the spinal cord activity maps estimated from these conditional models ( Fig 4A and 4B ) and that from the unconditional model ( Fig 3 , spinal level ) supports the idea that the observed learning-dependent modulation in the spinal cord activity is not a mere consequence of changes in descending inflow due to cerebral plasticity , but rather is suggestive of intrinsic plastic changes that occur at the level of the spinal cord . Altogether , the results of such analyses suggest that both brain and spinal cord may assume different aspects of behavioral variability during skill acquisition . To specifically test the latter hypothesis , we performed an analysis of variance using hierarchical regression models , which allowed us to estimate learning performance variability ( dependent variable ) using all possible combinations of cortical , subcortical , and spinal levels ( independent variables ) ( see Fig 3 ) . Hence , seven models were constructed: three based on the contribution of only one level , three using a combination of two from the three levels , and one including all three levels . The proportionate amount of performance speed variability explained by each model was assessed through an adjusted R-squared measure analysis ( S3 Table ) . Based on the adjusted R-squared values in this hierarchical set , we then built a Venn diagram ( Fig 4C ) to visualize the relative influence of these different central nervous system ( CNS ) levels and their overlap in explaining performance speed variability over the course of learning . As shown in Fig 4C , activity in the spinal cord and the brain accounted for nonoverlapping portions of variability . Specifically , the cervical cord accounted for 24% of total explained variability , of which 81% was linearly independent from the contribution of cortical and subcortical regions in capturing performance variability during motor sequence learning ( see S3 Table ) . This suggests that distinct neural mechanisms were responsible for the observed activity changes at the brain and spinal cord level . In order to explore whether the interaction in brain and spinal cord activity is context independent or can be altered by experience and/or learning condition , we assessed functional interaction between these structures over the course of learning using psychophysiological interaction ( PPI ) analyses ( Eq 1 ) [24] . An individual-specific spinal cord ROI , centered on the C7 cervical level ( see S3 Fig for the applied ROI’s mask ) , was selected as a seed region based on the main effect of practice during both CS and SS conditions ( Fig 2A ) . We evaluated the changes in functional connectivity between the spinal ROI and all the brain voxels , in proportion to the subjects’ improvement in performance speed during both CS and SS conditions . It is important to note that such analysis did not reveal any significant change in brain/spinal cord connectivity during the SS condition . By contrast , a cluster in the right primary sensorimotor cortices showed a significant decrease in positive correlation with the cervical cord in proportion to the amount of learning during the CS condition ( Fig 5A , right panel , contralateral M1 and S1 hand area , p < 0 . 01 , corrected for multiple comparisons using GRF ) . Furthermore , we observed an increase in negative correlation between the spinal cord ROI and a cluster mainly located in the left anterior cerebellum in proportion to the amount of learning during the CS ( Fig 5A , left panel , lobule IV and superior cerebellar peduncle , p < 0 . 01 , corrected ) . Finally , to investigate the amplitude and direction of functional connectivity at different stages of learning compared to baseline , we evaluated the linear correlation between BOLD signals of the reported brain areas and the spinal cord ROI in both early ( the first two blocks ) and late ( the last two blocks ) training periods during SS and CS conditions ( Fig 5B ) . This analysis confirmed the presence of training-induced changes in BOLD signal synchronization between the spinal cord and brain , which depended upon the acquisition of a new motor sequence ( repeated measures ANOVA , significant interaction between learning condition ( CS versus SS ) and time ( early versus late ) , p < 0 . 05 for M1/S1 , and p < 0 . 01 for cerebellum ) . Also , it shows that , as learning proceeds throughout CS condition , the activity within the cervical ROI becomes less correlated with that of primary sensorimotor cortex ( t22 = 2 . 8 , p < 0 . 05 , corrected ) but becomes more negatively synchronized with that of the anterior cerebellum ( t22 = 3 . 3 , p < 0 . 01 , corrected ) . In order to examine the possibility of nonspecific changes ( unrelated to the MSL process ) in motor output during the course of learning , we measured the power of electromyography ( EMG ) signals during training in both CS and SS conditions . A separate control group of subjects ( n = 10 , median age = 23 y; number of males = 5 ) was recruited and underwent the same motor learning procedure as the experimental group described above , except that testing was carried out in a magnetic resonance scanner simulator while recording EMG signals from several related extrinsic ( flexor digitorum superficialis and extensor digitorum ) and intrinsic ( first dorsal interosseous and flexor digiti minimi ) hand muscles . This allowed us to test for any motor execution-related differences at the periphery , which might be causing the pattern of spinal cord activity differences observed across conditions . The motor learning curves , as measured by speed performance in the control group , were very similar to that of the experimental group ( S5 Fig; mean duration of the first two blocks in the control group compared to the experimental group during the CS: p = 0 . 77 [two-sample t-statistics , df = 33] , and during SS: p = 0 . 57; and mean duration of the last two blocks during CS: p = 0 . 75 , and during SS: p = 0 . 89 ) . EMG analysis during the CS and SS conditions did not reveal any significant effect of condition on the normalized EMG power or root-mean-square ( RMS ) in any of the tested muscles ( main effect of learning condition: p > 0 . 14 in all comparisons; S6 Fig ) . Also , the normalized EMG power ( and RMS ) was not statistically different over the course of training between CS and SS conditions ( condition × block interaction: p > 0 . 2 , df = 11 ) . These results suggest that the observed plastic changes at the spinal cord level were not a by-product of variations in the applied muscle force ( as estimated by EMG power ) during motor learning .
Compared to brain functional neuroimaging studies , there is a scarcity of experiments that aimed to scan the activity of the spinal cord . This can be explained by the fact that imaging the spinal cord raises several challenges . These stem primarily from the small size of this structure , the magnetic field inhomogeneities caused by the many different tissue types surrounding it , and the movements induced by respiration [25 , 26] . Accordingly , in the present study , we undertook a series of steps , both during data acquisition as well as analyses , in order to overcome these challenges and to ensure that activations detected in the spinal cord were genuine and not the results of various artefacts . During data acquisition , we first used a slice prescription pulse sequence that yielded at least one good-quality slice per cervical level and minimized the inhomogeneity due to the surrounding tissue . Second , we acquired gradient-echo field map images to estimate field inhomogeneities . Third , we used an in-plane spatial resolution of 2 . 5 mm2 , hence allowing there to be at least 12 voxels covering the spinal cord while maintaining sufficient signal-to-noise ratio ( SNR ) . Furthermore , during data preprocessing , we first used the acquired field map images to correct for field distortions and carried out an independent component analysis to extract and remove the cardiac and respiratory-related physiological noise components from data based on their spatiotemporal characteristics . Second , we then estimated the spinal motion parameters ( three rotations and three translations ) and excluded from analysis the volumes in which the spinal cord displacement was higher than 1 mm . Accordingly , the number of deleted volumes ranged from 0 to 14 with a median of 2 per subject out of the total number of acquired volumes , which varied from 283 to 453 depending on participant’s execution speed ( mean of 384 . 2 and median of 386 volumes ) . Third , we used the motion parameters as confounds in the GLM analysis to account for motion artefacts . Other important arguments against the possibility that our results reflect motion artifacts are ( a ) the fact that block-related averages of the spinal BOLD signal ( Fig 2A , inlets ) , which are model free and not convolved by the hemodynamic response function ( HRF ) , showed the typical 8-s delay from the block onset until they reach the plateau and ( b ) that we also defined subject-specific functional masks that included two additional voxels outside the spinal cord in each direction , both along the anteroposterior and mediolateral axes ( S7 Fig ) , hence allowing the detection of spurious activity outside the spinal cord confines , if any . At the end , our results were robust , as all detected activity peaks were focal , within the spinal cord boundaries , and were located at the expected site of the spinal cord ( mostly ipsilateral and centered on C7–C8 ) , in line with the location of motoneurons innervating the finger muscles involved in the motor task . However , because of the large magnetic inhomogeneity produced by the intervertebral disks ( Fig 1C ) , our analysis is unable to determine whether the two activated clusters in the rostrocaudal plane at the C7 and C8 cervical levels ( Fig 3 ) are part of the same functional unit ( seemingly split because of the presence of a low SNR slice at the disk level ) or whether they actually represent two functionally distinct clusters . Yet , when lowering the statistical threshold to Z = 2 , we observed that the two clusters joined together , hence possibly supporting the former case . Despite this uncertainty , however , our results strongly suggest that spinal activity was not a mere reflection of the speed with which subjects executed the finger movements; even though movements in the SS condition were faster , the spatial extent and amplitude of the spinal activation were greater during the CS condition . Importantly , in counting the number of activated voxels , we did not perform any spatial smoothing in order to prevent the spread of activity from surrounding voxels , hence yielding independent measures of activation amplitude and spatial extent ( although the partial volume effects in BOLD fMRI acquisition can cause some potential confounds in this analysis ) . Lastly , spinal cord activity was modulated in association with behavioral measures of performance speed in the CS condition only ( and not the SS condition ) , hence supporting a process that is associated with both the complexity of the motor sequence and improvements in performance during learning . Because we were interested in acquiring data covering both the brain and spinal cord , the in-plane spatial resolution afforded in our present fMRI study was somewhat limited ( 2 . 5 x 2 . 5 mm2 in-plane ) . Considering the small size of the spinal cord , this makes it difficult to draw a clear-cut conclusion on the precise location of focal task-related activity within the cervical cord . Nevertheless , because of the larger size of the cervical cord along the mediolateral versus anteroposterior axis , we believe that the resolution used here allowed us to be relatively confident about the ipsilaterality of the focal activation , and less so about the dorsal/ventral localization of the activity . Accordingly , we would argue that the peaks of activity corresponding to the main effect of practice during both CS and SS conditions ( Fig 2A ) , as well as to the learning-related activity in the CS condition , were mostly located on the ipsilateral side of the cervical cord in line with the hand used during the motor task and the known spinal cord anatomy . Our study brings an important methodological contribution to the field of motor learning through the fact that the present pattern of functional motor learning-related cervical plasticity was observed for the first time using simultaneous brain/spinal cord fMRI in a large group of subjects . The innovative way of making use of the extensive field of view of the Siemens TIM Trio MR system ( 50 cm in the rostrocaudal plane ) in the current study allowed us to investigate quantitatively functional interactions between those structures and get insights into the neural substrates mediating motor sequence learning at all levels of the central nervous system . To acquire such data through imaging of both brain and spinal cord , some researchers have previously used custom-made coils [27] or dedicated imaging pulse sequences [28] . Opening the field for future studies , the present study employed regular equipment and sequences provided by the manufacturer , hence facilitating further the generalization of the imaging methods used here . Finally , contrary to most of the previous spinal imaging work that has reported activation maps at the subject level only , with the exception of a few reports on spinal cord modulation of pain [29 , 30] , the present study reports group data , because we developed an analysis procedure that allowed us to overcome several challenges , including those related to image normalization and alignment , a low SNR that limits the power of statistical analyses , and the physiological noise inherent to functional imaging of the spinal cord . A limitation of the current study is the extent to which motor learning can be decoupled from performance speed changes , given that the performance speed measure itself is an indicator of both learning and better motor execution . To address this issue , we used a control task ( SS ) , in which subjects knew the sequence very well , but performance speed was significantly increased over the course of training . Importantly , no modulation in the spinal cord activity was identified with respect to performance speed during this control task . Furthermore , the present pattern of results does not appear to be related to a possible confounding factor like differences in force used to perform the two tasks , as an independent study investigating EMG activity associated with learning did not reveal any significant disparities in EMG power ( measured by root-mean-squared values ) , which is a good estimate of applied muscle force during voluntary contractions . Finally , another means previously used to investigate the effects of learning , nonconfounded by unspecific changes due to task execution , has been to conduct functional connectivity analyses of resting state data [31] . Interestingly , this approach has recently been performed successfully using BOLD fMRI in the spinal cord [32 , 33] . However , the experimental design employed in the present study did not allow us to carry out such analyses because of the limited number of acquired volumes during rest periods between blocks of task-related activity . Thus , our results do not permit us to determine whether the task-based functional changes observed at the cervical level persisted after practicing the learning task , a finding that awaits further investigation . In line with early seminal work by Yoshizawa and colleagues [34] , there has been increased interest in recent years in investigating spinal cord BOLD activity during the execution of simple motor tasks [35–38] . While these studies have demonstrated the feasibility of BOLD spinal cord imaging during simple motor performance , none have examined the neural correlates of motor learning in the spinal cord or acquired simultaneous functional images of the brain and spine during performance of a motor learning task . Using another contrast mechanism based on proton density , called signal enhancement by extravascular water protons ( SEEP ) [39] , other researchers have also found strong motor- and sensory-related activity within the spinal cord . Recent studies comparing the BOLD and SEEP mechanisms have also consistently shown comparable results using either of these contrasts [37] . Yet , the present results expand our understanding of motor skill acquisition by showing that the spinal cord is an integral part of the neural network involved in this process . Until now , models of motor sequence learning in humans have elegantly described the experience-dependent plasticity that occurs in brain regions at different phases of the acquisition process but have not assigned any functional plasticity at the spinal cord level . At the cerebral level , the early stage of learning is supported by widespread activations predominantly in both the corticostriatal and corticocerebellar networks [1 , 2] . Our findings , however , show that in addition to these two systems , learning a new sequence of movements also includes plasticity within the cervical spinal cord . Furthermore , our results reveal the presence of intrinsic functional plasticity in the spinal cord that is associated with learning-related changes in motor performance and that is linearly independent from that of supraspinal structures . Although this does not exclude the possibility of a nonlinear relationship between activities in cervical and sensorimotor brain regions , it does eliminate the simplest possible alternative interpretation , which is that learning-related changes in cervical activity are merely mirroring the ongoing higher-level cerebral plasticity . Altogether , our results thus lend strong support to the idea that intrinsic plasticity can be induced at the spinal cord level in the early stages of motor learning . Although still conjectural , two possible reasons may explain the spinal cord’s active role during motor skill acquisition . First , together with learning-induced cortical and subcortical plasticity , the spinal cord could be another site where motor memories are stored . Previous studies have found that different types of motor memories can be encoded at the spinal level . For instance , animal research has shown that central pattern generators can be recovered in partially or totally transected cord following locomotor training [7] . In human studies with intact spinal cord , long-term changes in Hoffman reflex ( H-reflex ) in highly skilled individuals as compared to nonskilled individuals have also been reported , hence suggesting that the spinal cord is capable of encoding local motor memories [10 , 11] . A second reason may involve the stabilization and facilitation of movement execution through modification of muscle and joint stiffness as the motor skill is being acquired during practice . One possible way that such a process is achieved might be through elevated cocontraction of antagonistic muscles , which is known to occur during early stages of motor learning [40] . Indeed , muscle cocontraction levels can be directly adjusted via inhibitory mechanisms ( e . g . , presynaptic or disynaptic inhibition ) at the spinal cord level [10] . In line with this view , we have previously demonstrated that the H-reflex , a local measure of excitability of sensorimotor pathway within the spinal cord , was systematically diminished over the course of motor sequence learning , in which the reduction in reflex amplitude was greater compared to a control simple sequence or random movements [14] . It has been suggested that this persistent decrease in the H-reflex amplitude could be the result of increased presynaptic inhibition of the Ia afferent transmission to motoneurons [8 , 10] . Thus , the increased local cervical activity over the course of learning observed in the current study might be related to such synaptic mechanisms responsible for the reduction of the H-reflex at the spinal cord level . While our current findings cannot distinguish between these two functions of the spinal cord in motor learning , they certainly give support to the idea that spinal cord plays an active , rather than passive , role in this process . Furthermore , the scanning paradigm used in the present study offers a necessary tool for future investigations that specifically aim to parse out the roles of the brain and the spinal cord in various stages of motor learning . As verified by our EMG analysis , the motor task in the current study required both activations of extrinsic and intrinsic hand muscles , which are influenced by both indirect and direct corticomotoneuronal connections [41 , 42] . It is thus conceivable that motor learning could also be associated with functional changes in spinal circuitry involving both the interneurons as well as motoneurons , as illustrated by the fact that cervical activations detected in our study were not exclusively confined to the ventral horn of the spinal cord . Interestingly , connectivity analyses revealed that activity within the spinal cord was functionally more synchronized with the cerebellum , but less synchronous with primary sensorimotor cortical areas as learning progressed . Several possible mechanisms might explain the observed pattern of connectivity between the brain and spinal cord . First , the decreased interaction between the spinal cord and sensorimotor cortical areas might reflect the reduced one-to-one control of individual muscles by sensorimotor cortex late in the learning , through processes such as chunking [43] or the recruitment of muscle synergies [44] . In the case of muscle synergies , for example , various models have posited the existence of a module relying on spinal interneurons that generates a specific pattern of muscle activation [44] . Thus , as learning of sequential movements progresses , it is conceivable that the development of new muscle synergies leads to a greater local functional integration within the spinal cord and to a reduction in the one-to-one linear relationship with sensorimotor areas , hence resulting in a reduced linear functional connectivity between the spinal cord and sensorimotor cortex later in the acquisition process . A second possible mechanism could be that there was a shift in attentional focus as learning progressed [45] , which could then explain the functional decoupling between the sensorimotor cortex and spinal cord . In fact , the latter hypothesis is supported by the reported alteration of corticospinal excitability induced by attentional modulation during the performance of a motor task [46] . Finally , another probable mechanism could rely on corticospinal inhibitory processes . Although there are no direct long-range inhibitory corticospinal projections , the learning-related decrease in functional connectivity between these two structures might be the result of an increase in the weight of supraspinal projections onto the inhibitory spinal interneurons , which then project to ventral horn motoneurons [47] . Regardless of the mechanism of action , which cannot be directly tested with fMRI , the results of our functional connectivity analysis reveal that , at the end of learning , there is no significant correlation between cortical and spinal cord activities , indicating that the two structures become functionally desynchronized as learning progresses . By contrast , the increase in negative synchronization between activity in spinal cord and cerebellum could be more likely related to inhibitory mechanisms . Indeed , several electrophysiological studies have reported that cervical cord activity is inhibited when stimulating anterior medial neurons of the cerebellar cortex [48 , 49] . As the cerebellum is particularly involved in the coordination of rhythmic and oscillatory movements [50] , such increased interaction ( in the form of an inhibitory effect ) between the cerebellum and cervical cord over the course of motor sequence learning might reflect the underlying neural mechanisms responsible for the temporal control of rhythmic finger movements [51] . On the other hand , another possibility that might explain these results is that the cerebellum is implicated in controlling muscle and joint stiffness by adopting optimal strategies [52] , which in turn can be achieved through various spinal inhibitory mechanisms , as explained above . In sum , we believe that the simultaneous imaging and analysis of motor functions of the brain and spinal cord in this study can benefit the neuroscientific community , which has so far been divided in its investigation of the neural substrates mediating neuroplasticity in only one of these two structures . The present findings also have important clinical implications for the rehabilitation of patients with spinal cord injuries , as they demonstrate that this part of the central nervous system is much more plastic than it was assumed before . Altogether , the present findings support the view that the cervical spinal cord plays a critical role in our ability to acquire new motor skills , although the neurophysiological mechanisms underlying such neural plasticity await further investigations .
Twenty-five healthy young adults ( 14 females , 11 males; median age: 24 y ) participated in the present neuroimaging study . A separate group of healthy young adults ( n = 10 , 5 females; median age: 23 y ) participated in the EMG recording/motor learning experiment . All participants were right-handed , as determined by the Edinburgh Handedness Inventory . The study was approved by the joint research ethics committee of the Regroupement Neuroimagerie Québec at the Centre de recherche , Institut universitaire de gériatrie de Montréal , which follows the policies of the Canadian Tri-Council Research Ethics Policy Statement and the principles expressed in the Declaration of Helsinki . All participants gave their informed written consent . Subjects were tested using a version of the MSL task [53] , in which they were required to perform self-generated finger movements with their nondominant ( left ) hand as quickly and with as few errors as possible . Prior to the experiment , subjects practiced the sequential movements briefly ( up to three correct consecutive repetitions per sequence ) via an fMRI compatible button box . During the experiment , however , subjects laid supine in the scanner and executed the task following written instructions , which appeared on a screen visible via a mirror attached to the head coil . During “Rest , ” subjects had to rest with their eyes open for as long as the instruction appeared on the screen . When the instruction “Sequence” appeared , participants had to execute repeatedly a five-element motor sequence ( 4-1-3-2-4; complex sequence condition [CS]; where digits 1 to 4 correspond to the digits of the left hand from 1 [the index finger] to 4 [the pinkie] ) . Finally , when the word “Control” appeared on the screen , participants were required to execute a simple , four-element sequence repeatedly ( 4-3-2-1; simple sequence condition [SS] ) . In total , subjects were administered 24 blocks of 60 movements each ( corresponding to 12 or 15 repetitions of the complex and simple sequence per block , respectively ) , separated by rest periods lasting 15 s each . The two experimental conditions were split evenly across blocks ( i . e . , 12 blocks each ) and alternated in a pseudorandom fashion across blocks ( Fig 1B ) , with no more than two consecutive blocks in the same experimental condition . The total task duration varied between 11 min and 48 s to 18 min and 53 s , with a median of 16 min and 6 s , depending on participant’s speed of movement execution . Reaction time ( time elapsed between two consecutive key presses ) , block duration ( time to accomplish each block ) , and errors ( number of incorrect key presses in each block ) were recorded . Images were collected using a 3T whole-body Siemens TIM TRIO scanner with simultaneous detection via the four-channel neck , 12-channel head , and 24-channel spine coils . A structural volume was acquired in the sagittal plane using a magnetization-prepared rapid gradient echo ( MPRAGE ) sequence ( TR = 2 , 300 ms; TE = 3 . 31 ms; FoV = 320 × 320 mm; matrix size = 256 × 256; 160 slices , slice thickness = 1 . 3 mm , in-plane resolution = 1 . 25 × 1 . 25 mm ) . For functional acquisitions , an echo-planar imaging ( EPI ) gradient echo sequence was used with the following parameters: TR = 2 , 500 ms; TE = 30 ms; FA = 90°; FoV = 160 × 160 mm; matrix size = 64 × 64; slice thickness = 4 mm; in-plane resolution = 2 . 5 × 2 . 5 mm , parallel imaging with an accelerated factor of 2 and GRAPPA reconstruction . In total , 35–37 transversal slices per volume were acquired , as described in a previous study [54] . Coverage of the cervical cord of each subject was achieved by recording functional data from the 15 axial slices spanning the C1 to C7 cervical vertebrae ( corresponding to the C1 to C8 cervical spinal segments ) . Slices were spaced from 80% to 120% of the slice thickness in order to cover both the brain and the cervical spinal cord up to the first thoracic ( T1 ) segment , and they were placed at an angle that was perpendicular to the C4 vertebral segment in order to get the best in-plane coverage of the spinal cord ( Fig 1C ) . Slices were centered alternately at the midvertebral body level and at intervertebral disks [54] . The variation in the number of slices across subjects was due to the intersubject differences in gross spinal and brain anatomy . This particular slice prescription ensured that , despite individual anatomical variations , each cervical segment was covered using an axial slice passing through its center , hence making possible the precise coregistration of the functional data across participants . The number of acquired functional volumes was variable , depending on the participant’s speed during the task . Finally , dual echo field map images ( TE1 = 4 . 92 ms , TE2 = 7 . 38 ms ) were acquired to correct for the susceptibility-induced geometrical distortions in EPI data . Image processing was carried out using the FSL software package [55] and in-house programs developed in MATLAB . First , each functional volume for each subject was split into the brain and the cervical cord in the inferosuperior direction . The segmented brain functional images were first processed using regular preprocessing steps including motion correction , high-pass temporal filtering ( σ = 100 s ) , non-brain-tissue removal using the Brain Extraction Tool ( part of FSL ) , spatial smoothing ( 6 mm Gaussian kernel ) , and registration to the Montreal Neurological Institute ( MNI ) standard space . For each subject , changes in brain regional responses were estimated using a model including responses to the task practice conditions ( CS and SS ) and their linear modulations by performance speed ( negatively correlated with block duration ) . These regressors consisted of boxcars convolved with a double-gamma HRF . Six rotation and translation motion parameters were also included in the model as confounds . Modulation by performance speed identified regions where response amplitude changed as motor behavior became faster across blocks of practice . The subject-level regression coefficients and their covariance maps were then input to a group-level analysis , which used a mixed-effects general linear model ( Z > 2 . 3 , corrected family-wise error using Gaussian random field theory , cluster significance threshold of p < 0 . 01 ) . The preprocessing pipeline for analyzing the spinal cord functional image comprised the following: ( a ) creating a mask of the cervical cord by visual inspection—the mask included at least two additional voxels outside the spinal cord on the left , right , dorsal , and ventral sides ( S7 Fig ) ; ( b ) correcting for motion; ( c ) removing volumes with an absolute motion value > 1 mm from the target image; ( d ) utilizing GRE field map unwarping to correct for geometrical distortion ( S8 Fig ) ; and ( e ) applying high-pass temporal filtering ( cutoff period = 100 s ) . Two subjects were excluded from further analysis because of excessive movement during fMRI scanning . We generated two datasets with different spatial smoothing parameters ( 0 and 6 mm Gaussian kernel ) , which were used in two separate general linear models analyses in order to estimate task-based activity at the group level ( see below ) . Preprocessed data with no smoothing were used in the PPI analysis to calculate the BOLD signal average within the ROI , as explained below . Furthermore , smoothing was performed inside a mask within the cervical cord to ensure there was no infiltration of the surrounding structures’ signals into the cervical BOLD signal . Manual registration was performed in order to align EPI cervical slices of every subject to the EPI cervical image of one of the subjects , which was selected as the template . To do so , for each subject the center of cervical cord in each slice was manually marked and then shifted in the lateral and anteroposterior axes to match that of the template ( S8 Fig ) . Because of the subject-specific prescription of axial slices and slice gap-size adjustments , the cervical EPI slices were already aligned along the rostrocaudal axis across subjects . As physiological noise has a major impact in the detection of spinal cord BOLD signal [26] , we then used independent component analysis to identify and account for noise components at the cervical cord level [56] . Thirty components were extracted for each subject , which accounted for about 95% of the BOLD signal variability . The components that met the following criteria were considered as noise: ( i ) the component’s time series had more power at high frequencies ( more than 50% of power was associated with frequencies larger than 0 . 08 Hz [31]; task frequency was around 0 . 02 Hz ) and ( ii ) the component’s spatial map showed more activated voxels outside the spinal cord ( more than 50% of significantly activated voxels [Z > 2 . 3] were outside the spinal cord mask ) . Overall , 4 to 12 components met both criteria in every subject , and their time series were thus included as confound in the subject-level general linear model analysis . For each subject , the preprocessed cervical BOLD responses were estimated using a model that included the task practice conditions ( CS and SS ) and their linear modulations by performance speed ( normalized to Z-scores with a standard deviation of one and a mean of zero ) , as well as nuisance regressors comprising the six rotation and translation motion parameters , the time series of noise components extracted from the independent component analysis , and the average of white matter and CSF signals extracted from the brain as described previously [31] . By normalizing the performance speed regressors to Z-scores , their GLM coefficients in each condition became independent of the absolute differences in motor speed across conditions; they rather reflected the relationship between normalized performance variability within each condition and BOLD signal change in the spinal cord . The subject-level regression coefficients and their covariance maps were then input into a group-level analysis , which used a mixed-effects general linear model [55] . The corresponding Z-statistics maps for the contrasts of interest were generated and corrected for multiple comparisons using Gaussian random field theory ( minimum Z > 2 . 3; cluster significance threshold , p < 0 . 01 , corrected ) . Furthermore , in order to compare the spatial extent of cervical activation across conditions , we identified all of the activated voxels ( Z > 2 . 3; p < 0 . 01 uncorrected ) within the expected spinal levels ( C6–C8; segments wherein motoneurons innervating the finger muscles reside ) based on the subject-level activation maps in each condition . In this analysis , we used the preprocessed data that were not spatially smoothed ( smoothing kernel of zero ) to ensure that the number of activated voxels count was not confounded by the spread of activity to the neighboring voxels caused by spatial smoothing . We then calculated the percent BOLD signal change averaged over voxels that were activated in both CS and SS conditions ( conjunction map ) . In order to compare the amplitude of cervical activation across conditions , we evaluated the percent signal change in the conjunction map in each block of practice for each subject . A simple technique based on the notion of partial correlation [19 , 20] was used to account for the possible supraspinal contribution to the spinal BOLD response when modeling the learning-related effects . For each subject , the cervical BOLD responses were again estimated using a model that included the task practice conditions ( CS and SS ) and their linear modulations by performance speed ( learning-related regressors ) , but this time we included the time series of n conditioning brain areas x1 , x2 , … , xn as covariates . The design matrix also included the time series of nuisance signals described earlier . The learning-related regressors were orthogonalized with respect to all confound variables; this is mathematically equivalent to subtracting and removing mutual dependencies of the conditioning brain areas from the spinal cord activity when estimating the effects of learning [20] . Two conditional models were tested . In the first model , we incorporated , as confound , all the main brain areas that are known to send efferent information and thus that can influence spinal cord excitability [21–23] ( i . e . , contralateral M1 , dorsal and ventral premotor cortices , supplementary motor area , anterior cingulate , S1 , ipsilateral M1 and cerebellum ) . For each region , the peak of activity from the main effect of practice during both CS and SS conditions was extracted as seed voxel ( S1 Table ) . In the second model , we incorporated , as confound , all brain areas that showed sequence learning-dependent modulation . The seed voxels for this model were extracted from the peaks of activity using performance speed as a variable of interest ( i . e . , parametric modulator ) , when modeling brain BOLD response during CS condition ( including M1 , dorsal premotor , S1 , putamen , and lobule V–VI of the cerebellum; Fig 3 top panels; S1 Table ) . For each subject and seed voxel , we calculated the average of the BOLD signal in a standard spherical mask ( radius = 6 mm ) around the seed . In order to quantify the relative contribution of the spinal cord and the brain to behavioral improvements during motor sequence learning , we measured the explained variance in hierarchical general linear models . Each regression model estimated behavioral improvements ( as measured by performance speed convolved with HRF ) using a design matrix that included the time series of different brain and spinal cord areas . To obtain these time series , we selected all activation clusters , either in the brain or the spinal cord , which were modulated by the performance speed ( Fig 3; S2 Table ) , including the sensorimotor cortex ( cortical level ) , putamen and cerebellum ( subcortical level ) , and C7 cervical segment ( spinal level ) . We then used principal components analysis in combination with an adaptive pruning algorithm [57] to estimate the number of non-noise components in each cluster of interest . On average , nine components per cluster were extracted for each subject . We then constructed seven hierarchical regression models using different combinations of cortical , subcortical , and spinal levels ( only cortical , only subcortical , only spinal , cortical + subcortical , cortical + spinal , subcortical + spinal , and cortical + subcortical + spinal ) . For each model , we calculated the adjusted R-squared value ( adjusting for the number of predictors in each model ) , which indicates the proportionate amount of variation in the performance speed explained by the time series of components from different levels ( S3 Table ) . PPI analyses [24] were performed to test the functional connectivity of the brain with a reference spinal cord ROI , in proportion to performance speed changes during practice . New generalized linear models were constructed at the individual level , using several regressors to model brain activity at each voxel over time ( y ( t ) ) : y=[β1…β7][x1…x7]T+[β8…β15][c1…c8]T+ε ( 1 ) Four regressors represented the main effect of practice in each condition ( x1 and x2 ) and their modulation by performance speed ( x3 and x4 ) . The fifth regressor was the mean activity in the spinal cord ROI ( x5; physiological regressor , see below ) . The last two regressors of interest represented the interactions between each of the psychological ( modulation by speed performance during CS and SS ) and the physiological regressors ( i . e . , x6 = x3 . x5 and x7 = x4 . x5 ) . The psychological regressors were convolved with HRF . The design matrix also included movement parameters ( c1–c6 ) , as well as average CSF ( c7 ) and white matter signals ( c8 ) as confound . The reference spinal cord ROI was selected based on both anatomical ( expected cervical level ) and functional ( task-related ) constraints in the functional space of each subject , in order to attain high sensitivity by selecting individual-specific spinal cord activation maps . For each subject , all the activated voxels in the conjunction maps between CS and SS main effect of practice ( Z > 2 . 3; p < 0 . 01 uncorrected ) within the C6–C8 spinal levels ( see S3 Fig ) were selected as ROI . A significant PPI indicated a change in the strength of the functional connectivity between any reported brain area and the spinal cord ROI , which was related to performance speed changes during practice . To examine the direction of functional connectivity changes over the course of learning , we then evaluated the correlation between the BOLD signals of the reported brain areas and the spinal cord region of interest early ( the first two blocks ) and late ( the last two blocks ) during training in each of the SS and CS conditions . In a control experiment , ten young healthy adults were recruited and underwent the same motor learning protocol as the main experimental group . No brain imaging was performed , but in order to mimic the experimental environment and positioning of the subjects as much as possible , the experiment was done in a mock scanner at the Functional Neuroimaging Unit , Montreal . Subjects laid supine in the mock scanner and executed the task following written instructions , which appeared on a screen visible via a mirror . Similar to the experimental group , subjects performed 24 blocks of 60 movements each ( 12 in CS , 12 in SS , pseudorandom alternation ) , separated by 15-s rest period blocks . Additionally , EMG signals were recorded from two extrinsic ( flexor digitorum superficialis [FDS] , and extensor digitorum [ED] ) and two intrinsic ( first dorsal interosseous [FDI] , and flexor digiti minimi [FDM] ) finger muscles during the experiment . FDS and ED mainly function as the flexor and the extensor of the middle phalanges of the four fingers , respectively . FDI and FDM , on the other hand , function as the flexor/abductor of the index finger and the flexor of the little finger , respectively . EMG signal was sampled at 5 , 000 Hz . First , for each muscle , bias and linear trends were removed from the raw EMG signal . Then , to calculate EMG power , full wave rectification was applied , followed by low-pass filtering ( butterworth filter , fco = 20 Hz ) to obtain the EMG signal envelope . Also , RMS values were calculated on the detrended and high-pass filtered ( butterworth filter , fco = 10 Hz ) EMG signal . For each extrinsic muscle and each block , EMG power and RMS were averaged across the whole block , as FDS and ED were activated throughout the task block ( during all four finger movements ) . However , for intrinsic muscles , EMG power and RMS values were calculated and averaged over time intervals centered on the index and pinkie key presses ( 150 ms before and after each key press ) in each block , as FDI and FDM are mainly activated during the index and pinkie movements , respectively . To correct for differences in the number of index ( or pinkie ) key presses across conditions , mean RMS values were divided by the square root of the number of key presses in each condition . Also , in order to obtain comparable measures across subjects , for each block , EMG power and RMS values were normalized by the average power and RMS values over all blocks of SS condition , respectively . Results are shown as mean ± SEM . Based on previous studies of spinal cord and brain imaging , 25 subjects were deemed to yield sufficient power to detect changes within the sensorimotor network [54] . Data were checked for normality and equality of variance across conditions . Unless otherwise indicated , statistical significance was determined using repeated measures two-tailed t-tests ( when comparing two conditions ) or repeated measures ANOVAs ( when comparing more than two conditions ) . Results were considered to be significant at p < 0 . 05 . | When we acquire a new motor skill—for example , learning how to play a musical instrument—new synaptic connections are induced in a distributed network of brain areas . There is ample evidence from human neuroimaging studies for this high plasticity of the brain , but what about the spinal cord , the main link between the brain and the peripheral nervous system ? Literature on animal models has recently hinted that spinal cord neurons can learn during various conditioning paradigms . However , human learning models by tradition assume that the spinal cord acts as a passive relay of information from the cortex to the muscles . In this study , we simultaneously acquired functional images of both the brain and the cervical spinal cord through functional magnetic resonance imaging , and we provide evidence for local spinal cord plasticity during a well-studied motor learning task in humans . We also demonstrate a dynamic change in the interaction of the brain and spinal cord regions over the course of motor learning . The present findings have important clinical implications for rehabilitation of patients with spinal cord injuries , as they demonstrate that this part of the central nervous system is much more plastic than it was assumed before . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Simultaneous Brain–Cervical Cord fMRI Reveals Intrinsic Spinal Cord Plasticity during Motor Sequence Learning |
Clostridium difficile is the most common cause of antibiotic-associated nosocomial infection in the United States . C . difficile secretes two homologous toxins , TcdA and TcdB , which are responsible for the symptoms of C . difficile associated disease . The mechanism of toxin action includes an autoprocessing event where a cysteine protease domain ( CPD ) releases a glucosyltransferase domain ( GTD ) into the cytosol . The GTD acts to modify and inactivate Rho-family GTPases . The presumed importance of autoprocessing in toxicity , and the apparent specificity of the CPD active site make it , potentially , an attractive target for small molecule drug discovery . In the course of exploring this potential , we have discovered that both wild-type TcdB and TcdB mutants with impaired autoprocessing or glucosyltransferase activities are able to induce rapid , necrotic cell death in HeLa and Caco-2 epithelial cell lines . The concentrations required to induce this phenotype correlate with pathology in a porcine colonic explant model of epithelial damage . We conclude that autoprocessing and GTD release is not required for epithelial cell necrosis and that targeting the autoprocessing activity of TcdB for the development of novel therapeutics will not prevent the colonic tissue damage that occurs in C . difficile – associated disease .
Clostridium difficile is a gram-positive , spore-forming anaerobe that infects the colon and causes a range of gastrointestinal disorders including diarrhea , pseudomembranous colitis , and toxic megacolon [1] , [2] . This is a major healthcare concern as the number and severity of C . difficile-associated disease ( CDAD ) cases have increased dramatically in recent years [3] . Two large toxins , TcdA and TcdB ( 308 kDa and 270 kDa , respectively ) , are recognized as the main virulence factors of C . difficile [4] , [5] . The C-terminal portion of these toxins is responsible for delivering an N-terminal glucosyltransferase domain ( GTD ) into the host cell [6] , [7] . The GTD inactivates Rho family GTPases including Rho , Rac1 , and Cdc42 [8] , [9] . While there are numerous studies that report the effects of toxin-mediated glucosylation in cells , a consensus as to the conclusion of these reports , taken together , has been difficult due to differences in cell types , toxin concentrations , and assay methods . In addition , it appears that TcdA and TcdB can elicit different effects under similar conditions [10] , [11] . In all reports , both toxins can induce a cytopathic effect characterized by cell rounding . In many reports , these cells go on to die by apoptotic mechanisms , but the time course can be up to 48 hours [12]–[19] . It has been noted , however , that apoptosis cannot be detected in cells treated with higher concentrations of TcdB [20] . In at least one study , the absence of apoptosis in cells treated with TcdB has led to suggestions of a necrotic mechanism of cell death [21] . The mechanism of GTD delivery for TcdA and TcdB involves binding a host cell receptor [22] , [23] , uptake by endocytosis [24] , [25] , pH-dependent pore formation [26]–[28] , translocation across the endosomal membrane , host-factor dependent autoprocessing [29] , and release of the GTD into the host cell cytosol [30] . Release is thought to allow the GTD access to the Rho-family GTPases tethered to the plasma membrane surface . An N-terminal sub-domain within the GTD is thought to serve as a membrane localization domain [31] . The autoprocessing function of the toxins is mediated by a cysteine protease domain ( CPD ) that follows the N-terminal GTD [32] . Inositolphosphates , predominantly inositol hexakisphosphate ( InsP6 ) , have been identified as the host factors responsible for inducing autoprocessing [29] . The InsP6-bound structures of the TcdA and TcdB CPDs reveal a positively charged InsP6-binding pocket that is distinct from the catalytic active site [33] , [34] . InsP6 binding is thought to trigger conformational changes that permit the formation of the substrate-binding pocket and alignment of the catalytic residues [35] . The three catalytic amino acids Asp587 , His653 , and Cys698 ( TcdB sequence ) and the P1 substrate recognition site , Leu543 , have been shown to be important for in vitro processing activity by genetic mutation [32] . Mutation and chemical modification of these residues has also been shown to prevent activity in various cell based assays [29] , [32] , [34] , [36] , [37] . For this reason , TcdB autoprocessing activity and GTD release have been considered important in the toxin mechanism , an idea which suggests that the CPD could serve as a useful target for novel small molecule inhibitor discovery . The objective at the outset of this project was to conduct a high-throughput screen for small molecules that inhibit TcdB-mediated cell death . Our first step toward exploring this potential was to evaluate apoptotic and necrotic markers as cell death indicators . In observing a necrotic response to TcdB , we decided to specifically focus on the question of whether the assay would be able to detect inhibition of TcdB autoprocessing . We constructed mutant TcdB proteins with deficiencies in either the autoprocessing or glucosyltransferase activities and tested their effects on cell viability . Our unexpected observation that the mutants killed cells rapidly and at concentrations comparable to wild-type led us to investigate the role of autoprocessing and GTD release in cell death and cell rounding in greater detail . In this report , we provide evidence that epithelial cells and porcine colonic tissue challenged with TcdB undergo a rapid , necrotic cell death that is not dependent on autoprocessing and GTD release .
The objective at the outset of this project was to conduct a high-throughput screen for small molecules that inhibit TcdB-mediated cell death . Our first goal was , therefore , to establish conditions for an assay that was sensitive and homogeneous . HeLa cells were seeded into 384 well plates and treated with TcdB at multiple concentrations for varying lengths of time . Cells were then simultaneously assayed for caspase-3/7 activation and ATP levels using fluorescent and luminescent indicators , respectively . At all concentrations and time points tested , TcdB failed to activate caspase-3 and -7 , central regulators in apoptotic cell death ( Figure 1A ) . Conversely , staurosporine , a known inducer of apoptosis , triggered significant caspase-3/7 activation at a 5 hour time point . Since the result appeared to be in conflict with a previous report showing that TcdB-treatment of HeLa cells induced an increased rate of caspase-3 activity [18] , we performed additional experiments using lower toxin concentrations , a 48 hour time point , and TcdA . We did not observe caspase-3/7 activation in any of the cells treated with TcdB and only saw TcdA-induced caspase-3/7 activation when the toxin was applied at concentrations of 100 nM ( Figure S1A ) . While our initial experiments were performed with TcdB purified from a recombinant Bacillus megaterium expression system , we did not observe caspase-3/7 activation when we tested TcdB purified from C . difficile culture supernatants ( Figure S1B ) . Despite the lack of caspase-3/7 activation , the TcdB treatments had a significant impact on cellular ATP levels ( Figure 1B ) . Decreases in ATP were observed after only 2 . 5 hours in cells treated with 1 , 10 , and 100 nM TcdB suggesting that these cells were no longer viable . The effect is specific to TcdB , as TcdA only impacted the viability at concentrations of 100 nM at 24 hours ( Figure S2A ) . While lower concentrations of TcdB can induce cell death after a 48 hour application , the effect does not appear to be dose dependent at the 48 hour time point ( Figure S2A ) . In an attempt to correlate the viability indicators with cytopathic events , mock and TcdB treated cells were visualized by light microscopy . At concentrations of 10 pM , a characteristic cytopathic ( cell rounding ) effect was observed . In contrast , cells treated with 10 nM TcdB for 2 . 5 hours had completely lost their membrane integrity ( Figure 1C ) . The rapid loss of ATP and membrane integrity suggested that cells treated with nM concentrations of TcdB were dying by necrosis . To further test this hypothesis , we assessed the effect of TcdB on LDH and HMGB1 release . LDH release was apparent 2 . 5 hours after intoxication and at an increased level after 8 hours ( Figure 1D ) . Similar values for LDH release are observed when the cells are treated with TcdB from C . difficile supernatants ( Figure S2B ) . Notably , LDH release is only detectable at toxin concentrations above 0 . 1 nM , consistent with the cell death data obtained with an ATP indicator ( Figure 1B ) . HMGB1 is a nuclear protein that is released into the cytoplasm when the cell is dying by necrosis . We found that at 10 nM TcdB , HMGB1 was released into the cytoplasm after 1 hour ( Figure 1E ) . As a result of these studies , CellTiterGlo , the luminescent indicator of cellular ATP levels , was deemed the best indicator of cell viability for high throughput screening . The rapid loss of ATP and membrane integrity , the release of LDH and HMGB1 , and the lack of caspase-3/7 activation all suggest necrosis is the mechanism of TcdB-mediated death in HeLa cells . We next generated autoprocessing-deficient mutants that could be used as negative controls in a secondary assay that would allow us to select for molecules that inhibit the autoprocessing activity of the toxin . Single amino acid point mutations were made in the TcdB autoprocessing active site ( C698S , C698A , H653A , and D587N ) and the cleavage site ( L543A ) . Proteins were expressed in the B . megaterium expression system and purified to homogeneity . All mutants were tested for their in vitro autoprocessing activity ( Figure 2A ) . TcdB autoprocessing can be induced with the addition of 1 uM InsP6 , and the amount of processing increases as the concentration of InsP6 increases . At all concentrations of InsP6 , TcdB C698S , TcdB C698A , and TcdB H653A were completely inactive in autoprocessing , as detected by Coomassie-stained SDS PAGE ( Figure 2A ) and densitometry ( Figure 2B ) . TcdB D587N and TcdB L543A had residual cleavage activity , but were significantly cleavage-impaired . Cleavage of D587N was not induced until 100 uM InsP6 was added , and the amount of processed toxin was reduced . We next wanted to confirm that the mutants were also defective for autoprocessing in the context of the cell . HeLa cells were treated with wild-type TcdB or autoprocessing deficient TcdB mutants for 50 min , lysed , and probed by Western blot using an anti-TcdBGTD antibody . Free GTD was detected in cells treated with wild-type TcdB but was not detected in cells intoxicated with TcdB mutants ( Figure 2C ) . The same lysates were probed with an antibody specific for unglucosylated Rac1 . Rac1 is glucosylated even when the cells have been treated with autoprocessing mutants . These data suggest that in cells treated with TcdB autoprocessing mutants , the GTDs are being translocated into the cytosol , but they remain tethered to the endosome where glucosylation of Rac1 can still occur . To test the hypothesis that small molecule inhibitors of TcdB autoprocessing could be detected in a cell based screen , we assessed cell viability in response to three of the TcdB autoprocessing mutants: TcdB C698S , TcdB C698A , and TcdB L543A . HeLa cells were treated for 2 . 5 hours with multiple concentrations of TcdB and the TcdB mutants , and viability was assessed using CellTiterGlo . Unexpectedly , the autoprocessing deficient mutants were found to induce cell death at concentrations comparable to TcdB ( Figure 3A ) . To test whether this response was unique to HeLa cells , we performed similar experiments with Caco2 cells , an epithelial cell line derived from human colon . As with the HeLa cells , wild-type and autoprocessing deficient TcdB mutants induced a decrease in cellular ATP at similar concentrations in Caco2 cells ( Figure 3B ) . Caspase-3/7 activation was not detected in HeLa cells treated for 25 hours with autoprocessing deficient TcdB mutants ( Figure 3C ) , and the amount of LDH released in HeLa cells treated with wild-type TcdB and the TcdB C698S , C698A , and L543A autoprocessing mutants was equivalent ( Figure 3D ) . Finally , HeLa cells were treated with 10 nM wild-type and mutant TcdB proteins in the presence of a live/dead cell indicator and imaged every 10 minutes over a 2 hour time course . A representative movie of what we observed is included in the supplemental material ( Video S1 ) . The percentage of dead cells quantified over six fields suggests that the kinetics of cell death are identical for the four proteins ( Figure S3 ) . Collectively , these data suggest autoprocessing is not required for TcdB-mediated necrosis in epithelial cells . The idea that TcdB-induced necrosis did not require autoproteolytic release of the GTD suggested that the TcdB glucosyltransferase activity would also not be required for cytotoxicity . To test this hypothesis , single amino acid point mutations were made in the glucosyltransferase active site ( D270N , D270A , Y284A , W520A , and N384A ) based on the crystal structure of the TcdB GTD bound to UDP-glucose [38] . Proteins were expressed in the B . megaterium expression system and purified to homogeneity . All mutants were tested for their in vitro glucosyltransferase activity in the presence of purified Rac1 and UDP[14C]glucose , and all were impaired relative to wild-type ( Figure 4A ) . Of the five mutants , the TcdB D270N mutant showed the greatest defect in in vitro glucosyltransfer , with residual activity only evident in the highest concentrations of enzyme and substrate ( Figure 4B ) . Even with differences in the amount of residual activity , all five mutants were defective in the modification of Rac1 in cells ( Figure 4C ) . Furthermore , all 5 mutants were capable of inducing a cytotoxic effect similar to that of wild-type TcdB when applied to HeLa cells ( Figure 4D ) and Caco-2 cells ( data not shown ) . We interpret these data to mean that the TcdB cytotoxic effect does not require the glucosyltransferase activity of the toxin . The observation that TcdB autoprocessing mutants were able to glucosylate Rac1 in cells ( Figure 2C ) suggested that they would induce rearrangements in the actin cytoskeleton that result in the cytopathic ‘rounding’ phenotype . To investigate this , HeLa cells were treated with multiple concentrations of wild-type and mutant TcdB proteins and imaged every 10 minutes over a 2 hour time course . The percentage of round cells was quantified over six fields for each concentration and time point . At a 10 pM concentration , we observed similar rounding kinetics for TcdB and the three TcdB autoprocessing-deficient mutants ( Figure 5A ) . Differences in the kinetics of rounding began to appear at a concentration of 100 fM ( Figure 5B ) but were not fully evident until the concentration of toxins was dropped to 1 fM ( Figure 5C ) . The full dataset collected at concentrations spanning 8 orders of magnitude and a movie of what we observed with 10 fM wild-type TcdB is included in the supplemental material ( Figure S4 and Video S2 ) . While not required for cytotoxicity , autoprocessing and GTD release appear to be important for cytopathic processes that occur at very low concentrations . In HeLa cells , we see that at concentrations where cytopathic effects can be observed ( 1 fM–10 pM , Figure 5 ) , the cells are not dead ( Figure 3A ) . These data provide a clear distinction between the cytotoxic and cytopathic effects induced by TcdB . The distinction between cytopathic and cytotoxic events in cell culture led us to question if either event might correlate with disease pathology . Since the formation of necrotic lesions in the colon is a hallmark of CDAD pathology , we sought to determine the concentration of toxin required to induce these effects and whether autoprocessing was required . Porcine colonic explants were incubated with multiple concentrations of toxin for 5 hours . The tissue was fixed with formalin , embedded in paraffin , and sections were stained with H&E ( Figure 6A ) . The slides were scored in a blinded fashion and given a score ( 0–3 ) to reflect the level of epithelial damage ( Figure 6B ) . Damage ranged from a mostly intact surface epithelium to mucosal loss of 50% or greater in the depth of colonic crypts . The scores indicated a loss of surface epithelium in tissue treated for 5 hours with 10 nM TcdB and TcdB C698A . There was little damage in tissues treated with a buffer control or in tissues treated with wild-type TcdB and TcdB C698A at a concentration of 10 pM . Statistical analysis by two-way ANOVA revealed a significant difference in scores for tissues treated with the toxins over the range of concentrations ( p<0 . 001 ) , while there was no statistical difference between tissues treated with wild-type TcdB and TcdB C698A . A subsequent Bonferroni's test revealed that scores given to tissue treated with 10 nM TcdB and 10 nM TcdB C698A were significantly different from scores given to tissue treated with 10 pM TcdB and 10 pM TcdB C698A ( p<0 . 001 ) . The tissues were stained with an anti-pan cytokeratin antibody to confirm the keratin positive cells at the luminal surface of the colon were disrupted ( Figure 6C ) and an anti-activated caspase-3 antibody to confirm that the toxin treatment did not induce an apoptotic response ( Figure 6D ) . The data reveal a correlation between the concentration of toxin required to kill epithelial cells in culture with the concentration required to disrupt epithelial integrity in colonic tissue and indicate that autoprocessing is not required for tissue damage .
TcdB is a multi-functional protein with a central role in CDAD pathogenesis . Our goal at the outset of this study was to conduct a screen for small molecule inhibitors that could aid in the dissection of the TcdB mechanism and the generation of new leads for therapeutic intervention . Our strategy was to combine a cell-based phenotypic screen with target-specific secondary assays . In the course of setting up our screening assays , we made two unexpected observations that warranted further investigation . First , in contrast to a previous report [18] , TcdB did not trigger the induction of apoptosis in cultured epithelial cells as measured by caspase-3/7 activation ( Figure 1A , S1 ) . Since there was an overlap in the cells , concentration of toxin , and timepoints used for analysis , we are left to speculate that the difference stems from advances in the detection reagent . The newer reagent for detecting caspase-3/7 activation allows one to directly quantitate the relative quantity of activated caspase-3/7 as opposed to the overall rate of caspase activity . While TcdB-treatment did not induce the activation of caspase-3/7 , the rapid ATP depletion observed in both HeLa ( Figure 1B , 3A , S2A ) and Caco2 ( Figure 3B ) cells suggested that the mechanism of TcdB-induced cell death was likely necrosis . The observed loss of membrane integrity ( Figure 1C ) , rapid LDH ( Figure 1D , S2B ) , and HMGB1 release ( Figure 1E ) support this conclusion . We next questioned whether a cell-based assay for small molecule inhibitors of TcdB-induced necrosis would allow us to detect molecules that interfered with autoprocessing . We were particularly interested in targeting the autoprocessing activity of the toxin since , in theory , one could identify molecules that either activate ( e . g . InsP6 ) or inhibit the function of the cysteine protease domain . We generated five TcdB point mutants in which key residues of the cysteine protease active site or cleavage site were mutated . Three of these mutations , C698S , C698A , and L543A , rendered TcdB non-functional for InsP6-induced autoprocessing in an in vitro assay , even when InsP6 was added at a 1 mM concentration ( Figure 2A , 2B ) . The mutants were also defective for autoprocessing in the context of cells since free GTD could be detected in cells treated with wild-type TcdB but not in cells treated with the autoprocessing mutants ( Figure 2C ) . While we cannot rule out the possibility of an alternate cleavage mechanism that results in a quantity of free GTD that is less than the detection limit of the assay , the free GTD concentration generated from such a mechanism would be too small to account for the identical cytotoxicity profiles observed in Figures 3A and 3B . The unexpected observation that cytotoxicity does not require autoproteolytic release of the GTD led us to directly test whether the glucosyltransferase activity of the toxin was required ( Figure 4 ) . We generated five single amino acid point mutants of TcdB that differed in their residual glucosyltransferase activities in vitro ( Figure 4A , 4B ) . Despite the different enzyme activity levels , all were significantly impaired relative to wild-type TcdB in their capacity to modify Rac1 in cells ( Figure 4C ) , and all were comparable to wild-type TcdB in their cytotoxic effects ( Figure 4D ) . These data are consistent with the observation that autoprocessing is not required and suggest that the cytotoxic response to TcdB is triggered by an event upstream of GTD release . While not required for cytotoxicity , autoprocessing and GTD release are important for cytopathic processes that occur at low concentrations [29] , [32] , [34] , [36] , [37] . Our data are consistent with these previous reports and indicate differences in rounding kinetics emerging at concentrations of 100 fM ( Figure 5C and SF4 ) . While our Western experiment indicated TcdB autoprocessing mutants were still able to modify Rac1 in cells ( Figure 2C ) , a similar observation has been made for a non-cleavable form of TcdA and is thought to reflect continuous vesicle trafficking and an exchange of membranous compartments that allow the uncleaved toxin to come into contact with the membrane-bound GTPases [39] . This capacity to modify Rac1 while still tethered to the endosomal membrane presumably accounts for the similar rounding kinetics that we observed when the TcdB autoprocessing mutants were applied to HeLa cells at concentrations of 1 pM and higher ( Figure 5A , S4 ) . The concentrations of TcdB needed to induce cytopathic effects ( ≤1 fM , Figure S4 ) are significantly lower than what is required to induce the cytotoxic effect ( 1 nM , Figure 3 ) . At a concentration of 10 pM TcdB , the cells are clearly round ( Figure 5A ) but not dead ( Figure 3 ) . The distinction between cytopathic and cytotoxic events in cell culture raises the question of whether either process correlates with mechanisms of pathology observed in the host . To address this question , we decided to test what concentration of toxin was required to induce epithelial cell damage in colonic tissue explants . Visual assessment of H&E stained colonic tissue integrity in a blinded fashion indicated damage with treatments of 10 nM TcdB but not with 10 pM TcdB ( Figure 6 ) . Similar observations were made with the TcdB C698A mutant suggesting that the damage that occurs to colonic tissue in response to TcdB does not depend on the autoprocessing activity . Pan-cytokeratin staining confirmed that the cells on the luminal surface of the tissue remained intact in the presence of 10 pM TcdB or TcdB C698A but were being disrupted in samples treated with 10 nM TcdB , 10 nM TcdB C698A , or 100 uM staurosporine . The staurosporine control revealed strong caspase-3 activation into the crypts ( Figure 6D ) . The untreated control tissue demonstrated a low level of caspase-3 activation in the cells on the luminal surface and strong activation in single cells coming off the surface of the tissue . Tissues treated with 10 pM TcdB and TcdB C698A showed caspase-3 activation levels similar to those of the untreated tissue . Tissue treated with 10 nM TcdB or TcdB C698A demonstrated even lower levels of caspase-3 activation , presumably because the cells on the luminal surface have been shed . Unlike the untreated , staurosporine-treated , and 10 pM TcdB-treated tissues , caspase-3 activation was generally not observed in the cells that were in the process of being shed in tissues treated with 10 nM TcdB or TcdB C698A ( Figure 6D ) . This suggests that tissue damage is not only independent of autoprocessing activity , but also not likely due to apoptosis . The phenotypic differences with concentration led us to wonder what concentration of toxin is present in the colons of individuals experiencing the symptoms of CDAD . We found only one published report , where TcdB was quantitated using a real-time cell analysis system [40] . In this report , the TcdB concentrations in stool samples from 10 patients experiencing mild to severe symptoms of CDAD ranged from 4 . 9 pM to 413 pM with a mean concentration of 146 pM . Presumably , the concentration of TcdB would be much higher at the colonic epithelium prior to dilution by diarrhea . Of note , the average TcdB concentration in samples from 9 individuals who were not experiencing CDAD symptoms was 1 pM , with a range of 0 . 1 pM to 3 . 3 pM . This analysis suggests that the cytotoxic effects observed in cells and tissues treated with 1 to 10 nM TcdB are better correlated with pathology than the cytopathic effects that are induced at 1 fM concentrations . Our data suggest that inhibiting TcdB autoprocessing will not prevent the colonic tissue damage observed in C . difficile associated diseases . However , while the colonic epithelium is the primary barrier separating C . difficile from the host , it is possible that the autoprocessing function of TcdB is important in another setting relevant to pathogenesis . For example , the colonic explant model used in this study does not account for the impact of the toxins on inflammation or the potential impact of an anaerobic environment . Evaluating the effect of autoprocessing- and glucosyltransferase-deficient toxins in an animal model of C . difficile infection therefore represents a priority for future studies . In addition , it will be important to define the mechanism of TcdB-mediated necrosis in cells and tissue . Relevant comparisons may come from the study of other toxins . For example , the Bordetella pertussis adenylate cyclase ( AC ) toxin is known to have multiple mechanisms that contribute to cytotoxicity [41] . Identifying the autoprocessing- and glucosyltransferase-dependent and –independent aspects of TcdB-mediated pathology represents an exciting path for future study .
This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . Animal husbandry and experimental procedures related to the porcine colonic explants were performed in accordance with the Vanderbilt University Institutional Animal Care and Use Committee ( IACUC ) policy . Discarded colon tissues were obtained from pigs following euthanization at the end of IACUC-approved animal use protocols . Animal husbandry and experimental procedures related to the generation of the anti-TcdBGTD monoclonal antibody were performed in accordance with the Washington University Animal Studies Committee policy , approval number 20100113 . Single amino acid point mutations were made in the TcdB autoprocessing active site ( C698S , C698A , H653A , and D587N ) , the cleavage site ( L543A ) , and the glucosyltransferase domain ( D270N , D270A , Y284A , N384A , and W520A ) using the QuickChange mutagenesis protocol ( Stratagene ) . The template for mutagenesis and clone for the production of wild-type TcdB was a B . megaterium expression vector encoding the strain 10643 of TcdB [42] . A similar clone was used for expression of recombinant TcdA [42] . Plasmids for expressing TcdA , TcdB , and TcdB point mutants were transformed into B . megaterium according to the manufacturer's instructions ( MoBiTec , Göttingen , Germany ) . 1 L of LB was inoculated with 35 mL overnight culture and 10 mg/L tetracycline and grown at 37°C and 230 rpm . At an OD600 of 0 . 3 , expression was induced with 5 g of D-xylose . Cells were harvested after 4 h by centrifugation and resuspended in 20 mM Tris , pH 8 . 0 , 500 mM NaCl and protease inhibitors . Cells were lysed by French press , and lysates were centrifuged at 48 , 000 g for 25 min . The proteins were purified by Ni-affinity chromatography , Q-sepharose anion exchange chromatography , and gel filtration chromatography in 20 mM HEPES , pH 6 . 9 , 50 mM NaCl . Proteins were expressed and purified as previously described [1] . HeLa and Caco2 cells ( cultured in DMEM , 10% FBS , 5% CO2 and MEM , 10% FBS , 5% CO2 , respectively ) were seeded in a black 384-well plate at a concentration of 3 , 000 or 1 , 000 cells/well , respectively . HeLa cells were intoxicated the next day , and Caco2 cells were intoxicated 36 h later . After intoxication , the cells were incubated at 37°C , 5% CO2 for either 2 . 5 h ( HeLa ) or 18 h ( Caco2 ) . The amount of ATP ( cell viability ) was assessed with a luminescence-based indicator , CellTiterGlo ( Promega ) . LDH release was assessed with a luminescence-based indicator , CytoToxGlo ( Promega ) . Caspase-3/7 activation was determined using a fluorescent indicator , Apo-One ( Promega ) . Staurosporine ( Sigma , 1 mM ) was used as a positive control for caspase-3/7 activation . Plates were read in a Biotek Synergy 4 plate reader . HeLa cells were seeded into a tissue culture treated chamber slide at 2×104 cells per well and incubated overnight . Cells were synchronized at 4°C and intoxicated with 10 nM TcdB for 1 h . Cells were then shifted to 37°C for 1 h . Media was removed from the cells , and the cells were washed with PBS . They were fixed with 4% paraformaldehyde at room temperature for 10 minutes and quenched with 1 mM glycine . Cells were permeated with 0 . 2% Triton X-100 in PBS for 5 minutes , washed in PBS , and blocked for 30 minutes in PBS , 2% BSA , 0 . 1% Tween 20 . Cells were stained with a monoclonal antibody against HMGBI ( Abcam , ab77302 ) , and an Alexa Fluor 488 anti-mouse antibody ( Invitrogen , A11001 ) . Cells were visualized with an LSM 510 Confocal microscope . 1 uL InsP6 stock solution ( 100× ) or buffer was added to 200 nM TcdB or TcdB autoprocessing mutant and incubated for 2 h at 37°C . The reactions were stopped with the addition of loading buffer and boiling and analyzed by Coomassie stained SDS PAGE . Genomic DNA of C . difficile clinical isolate 630 was obtained from American Type Culture Collection , and the region encoding residues 1 to 549 of TcdB , which is known to encode the substrate binding and enzymatic domains of the toxin , was amplified in frame with a carboxy-terminal ( His ) 6-tag using upstream primer:5′- CCGGATGTACAGTTGAGGGGGTAAAATGAGTTTAGTTAATAGAAAACAGTTAG -3′ and downstream primer 5′- GGTCCTCAATGATGGTGATGGTGATGAAGATTATCATCTTCACCAAGAGAACC -3′ . The resulting product was cloned into plasmid pcDNA3 . 1 ( Invitrogen , Carlsbad CA ) and sequenced to ensure fidelity of the amplified product . The gene was then released with restriction enzymes BsrG1 and AgeI and cloned into similarly digested vector pHIS1525 ( MoBiTec ) , placing the gene under control of a xylose-inducible promoter . Recombinant protein was expressed in B . megaterium and purified by sequential nickel affinity and gel filtration chromatography . Two mice were immunized bi-weekly by intraperitoneal injection with 100 µg purified TcdB-GTD . Three days after the third vaccination , splenocytes were harvested and fused to P3X63Ag8 . 6 . 5 . 3 myeloma cells using polyethylene glycol 1500 [43] . Hybridomas producing anti-TcdB-GTD MAbs were identified by ELISA , subcloned by limiting dilution , and purified by protein G immunoaffinity chromatography . HeLa cells were synchronized by cooling to 4°C and then intoxicated with 10 nM TcdB , autoprocessing mutant , or buffer . The cells were returned to 4°C for 1 h , and then shifted to 37°C for 50 min . The cells were harvested and lysed , samples were boiled , and proteins were separated by SDS PAGE . Samples were analyzed by Western with primary antibodies specific for the TcdB GTD , unglucosylated Rac1 ( BD , 610650 ) , total Rac1 ( Millipore , clone 23A8 ) , and GAPDH ( Santa Cruz Biotechnology , sc-25778 ) . Binding of an anti-mouse , HRP-conjugated secondary antibody ( Jackson ImmunoResearch Laboratories , 115-035-174 ) was detected with a LumiGLO kit ( Cell Signaling ) according to manufacturer's instructions . Unless otherwise noted , 100 nM TcdB or TcdB glucosyltransferase mutants and 2 uM Rac1 were mixed with 20 mM UDP-[14C]glucose ( 250 mCi/mmol , Perkin Elmer ) in a total reaction volume of 10 uL . The buffer contained 50 mM HEPES pH 7 . 5 , 100 mM KCl , 1 mM MnCl2 , 2 mM MgCl2 , and 0 . 1 mg/mL BSA . Reactions were incubated at 37°C for 1 h and stopped with the addition of loading buffer and boiling . Proteins were separated by SDS PAGE , and glucosylation of Rac1 was detected by phosphorimaging . HeLa cells were seeded in a black 96-well imaging plate ( PerkinElmer ) and incubated overnight . Cells were pretreated with live/dead cell imaging dyes ( Molecular Probes , R37601 ) and then treated with multiple concentrations of wild-type and mutant TcdB proteins . Cells were imaged in an environment-controlled chamber ( 37°C , 5% CO2 ) every 10 minutes over a 2 hour timecourse using an Opera High-Throughput Confocal Screening Microscope and Peltier-cooled , confocal CCD cameras . The percentage of dead cells and round cells was quantified over six fields for each concentration and time point using the Columbus Analysis software . Dead cells were defined as red cells with an intensity greater than 450 relative units , and round cells were defined as having an area less than 500 um2 and a width-to-length ratio of less than 0 . 4 . Colonic tissue was harvested from purpose-bred 25–35 kg , male or female , York-Landrace crossbred pigs . Following an overnight fast and immediately after euthanasia , a midline incision was performed and 15 cm of distal colon proximal to the rectum was excised and placed in PBS . The colon was opened , the luminal side was washed 3×5 min in 1 mM DTT to remove the mucus , and 3×5 min in PBS prior to dissection . Individual tissue sections were placed in wells of a 24-well plate . A nutrient buffer [44] containing ( mM/liter ) : 122 . 0 NaCl , 2 . 0 CaCl2 , 1 . 3 MgSO4 , 5 . 0 KCl , 20 . 0 glucose , 25 . 0 NaHCO3 ( pH 7 . 5 ) was pre-conditioned with HeLa cells overnight at 37°C and used to dilute the toxins . Explants were treated with wild-type TcdB , mutant TcdB , staurosporine ( 100 uM , Enzo Life Sciences , ALX-380-014-C250 ) or nutrient buffer for 5 hours at 37°C . The tissues were fixed with formalin for 56 h , washed in PBS , and transferred to cassettes . The tissue blocks were then embedded in paraffin , and 4 µm sections were cut and stained with hematoxylin and eosin ( H&E ) by the Vanderbilt University Translational Pathology Shared Resource core . Stained sections were coded and evaluated by six individuals , using a semi-quantitative injury scale: 0- no damage; 1-superficial damage , damage limited to intact surface epithelial cells; 2-loss of up to 50% of surface epithelial cells or gland length , crypts intact; 3-loss of over 50% of surface epithelial cells and damage in greater than 50% of gland length . An injury score was calculated as the mean score for sections evaluated seven times by six individuals . Statistical analysis was performed using a two-way ANOVA and Bonferroni's test . For keratin and caspase staining , sections were de-paraffinized with Histo-clear ( National Diagnostics ) and antigens were retrieved by citric acid . The sections were blocked with Serum-free protein block ( Dako ) , stained with a rabbit anti-pan cytokeratin or anti-active caspase-3 antibody ( Santa Cruz Biotechnology , sc-15367; Abcam , ab13847 ) , and diluted in Dako's antigen diluent with background reducing components overnight at 4°C . The sections were washed with PBS and incubated for 1 hr at RT with an AlexaFluor 546 donkey anti-rabbit antibody ( Invitrogen A10040 ) . The sections were washed with PBS and mounted with Prolong Gold with DAPI ( Invitrogen ) . H&E , pan-cytokeratin , and caspase-3 stained sections were imaged using an Ariol SL-50 ( Epithelial Biology Center Imaging Core ) . | Clostridium difficile is an anaerobic spore-forming bacterium that infects the human colon and causes diarrhea , pseudomembranous colitis , and toxic megacolon . Most people that develop disease symptoms have undergone antibiotic treatment , which alters the normal gut flora and allows C . difficile to flourish . C . difficile secretes two toxins , TcdA and TcdB , that are responsible for the fluid secretion , inflammation , and colonic tissue damage associated with disease . The emergence of hypervirulent strains of C . difficile that are linked to increased morbidity and mortality highlights the need for new therapeutic strategies . One strategy is to inhibit the function of the toxins , thereby decreasing damage to the colon while the patient clears the infection with antibiotics . Toxin function is thought to depend on an autoprocessing event that releases a catalytic ‘effector’ portion of the toxin into the host cell . In the course of trying to identify small molecules that would inhibit such a function , we found that TcdB induces a rapid necrosis in epithelial cells that is not dependent on autoprocessing . The physiological relevance of this observation is confirmed in colonic explants and suggests that inhibiting TcdB autoprocessing will not prevent the colonic tissue damage observed in C . difficile associated diseases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens",
"pathogenesis"
] | 2012 | Clostridium difficile Toxin B Causes Epithelial Cell Necrosis through an Autoprocessing-Independent Mechanism |
Heart valve anomalies are some of the most common congenital heart defects , yet neither the genetic nor the epigenetic forces guiding heart valve development are well understood . When functioning normally , mature heart valves prevent intracardiac retrograde blood flow; before valves develop , there is considerable regurgitation , resulting in reversing ( or oscillatory ) flows between the atrium and ventricle . As reversing flows are particularly strong stimuli to endothelial cells in culture , an attractive hypothesis is that heart valves form as a developmental response to retrograde blood flows through the maturing heart . Here , we exploit the relationship between oscillatory flow and heart rate to manipulate the amount of retrograde flow in the atrioventricular ( AV ) canal before and during valvulogenesis , and find that this leads to arrested valve growth . Using this manipulation , we determined that klf2a is normally expressed in the valve precursors in response to reversing flows , and is dramatically reduced by treatments that decrease such flows . Experimentally knocking down the expression of this shear-responsive gene with morpholine antisense oligonucleotides ( MOs ) results in dysfunctional valves . Thus , klf2a expression appears to be necessary for normal valve formation . This , together with its dependence on intracardiac hemodynamic forces , makes klf2a expression an early and reliable indicator of proper valve development . Together , these results demonstrate a critical role for reversing flows during valvulogenesis and show how relatively subtle perturbations of normal hemodynamic patterns can lead to both major alterations in gene expression and severe valve dysgenesis .
Formation of valves is a critical step in the development of a functionally mature heart , yet little is known about the mechanisms that initiate valve formation in vivo . In vertebrates , valves form from the endothelial cell layer located at the border between the atrium and the ventricle [1]–[3] . In fish , this region is called atrioventricular ( AV ) canal [1] , [4] , [5] and defines the endothelial ring [6] . The expression of genes specific to this territory depends on the activity of molecules secreted in the subjacent AV myocardium and on an elaborate combination of signaling pathways between the two cell layers , including Wnt/β-catenin , bone morphogenetic protein ( BMP ) , and Notch signaling [4] , [7]–[10] . Not surprisingly , aberrant patterning of the myocardial layer of the early heart can lead to valve defects , as the specification of the AV canal is impaired [6] . The analysis of zebrafish mutants has led to the identification of several cellular changes happening in the endothelial cell precursors during the process of valvulogenesis [1] , and it has been shown that some of these changes are associated with physical stimuli provided by blood flow [11] . Interestingly , valve morphogenesis is clearly dependant on the geometry of the beating heart chambers , further suggesting that the physical environment near the developing valves plays a critical role for their development [11] . Along with previous observations demonstrating the importance of intracardiac fluid flow for cardiogenesis [4] , [12]–[14] , this offers the exciting possibility that the genetic programs that govern valve formation in vivo depend on intracardiac hemodynamics . Harvesting this possibility has been challenging , as some attempts to uncouple contractility and flow have been taken to suggest that they play opposing roles in modulating cell shape within the developing heart [12]; other studies have suggested that flow forces regulate looping , cell size and shape in the heart chambers , and the formation of trabeculae [5] , [12]–[14] . A recent publication highlights the uncertainty concerning the role of flow during heart valve development , since it reports a cardiac contractility mutant that can form normal valves [15]; thus , something more than the mere presence ( or absence ) of flow or contractility must be involved in directing valve development . The predominant model to explain endothelial cell response to flow envisions that the shear stress , which directly depends on the viscosity and the velocity of the blood , is the main physical stimulus . More recently , disturbed flow has been used as a general term to group abnormal flow patterns ( including low flow , oscillatory flow , flow separation , gradients , turbulence , and reversing flows ) , potentially leading to atherogenic stimulus for endothelial cells [16]–[18] . This hypothesis is indirectly supported by observation that , in vitro , endothelial cells can be responsive to disturbed flows [19] , [20] , leading to an atherogenic-like cell response [21] . Thus , an attractive hypothesis is that heart valves form as a developmental response to disturbed blood flows . A key prediction from this model is that altering flow patterns within the heartbeat cycle should directly affect valvulogenesis . In vitro approaches have so far been unsuccessful in addressing this question , possibly due to the absence of specific valve markers usable in vitro and to the difficulty to mimic in vitro the complexity of flow patterns observed in vivo . To circumvent these limitations , we characterized embryonic zebrafish heart flow in vivo to identify a critical feature of the flow pattern associated with valve specification and tested its importance using a set of experimental manipulations including both genetic and pharmacological approaches . Taking advantage of high-speed imaging , we quantified the flow patterns generated in the beating heart and compared them with anatomical landmarks of the heart specified by expression patterns of known genes . Using antisense morpholine oligonucleotides ( MOs ) and drugs to alter these flow patterns in zebrafish , we show that reversing flow is essential to trigger flow-responsive genes in the AV canal and for initiating valvulogenesis . Our findings validate a key prediction of a specific and local role for reversing flows during cardiogenesis .
In order to better understand the roles played by blood flow in heart and valve development , we have developed imaging techniques to capture cardiac motion and analyze blood flow . Imaging with these tools reveals dramatic changes in intracardiac blood flow patterns during cardiac development: as the heart enlarges , blood flow becomes increasingly bidirectional until the stage at which functional valve leaflets emerge at the boundary between the atrium and the ventricle ( Figure S1 , Video S1 , and unpublished data , see also [11] , [22] ) . Although reversing blood flows are at times visible in the atrium and ventricle , reversing flows are most pronounced at the AV canal in the second and third days of development ( Figure 1A–1C , Video S2 ) . We quantified the degree of reversing flow by measuring the fraction of the cardiac cycle during which there is retrograde flow , and term this the retrograde flow fraction ( RFF ) . RFF is largest at the AV canal at embryonic stages that precede valve formation . Our ability to observe intracardiac blood flow simultaneously with heart pumping dynamics and morphogenetic changes provides a direct means to assess the proposal that the presence of particular patterns of intracardiac blood flow play a critical role in heart valve development . To better understand how reversing flow relates to valve development at the molecular level , we analyzed the expression pattern by in situ hybridization ( ISH ) of three known shear-related genes at the AV canal: notch1b , a zebrafish Notch homolog [23]–[25] , klf2a , a transcription factor from the Kruppel-like factor ( Klf ) family [26] , and bmp4 , a secreted growth factor of the bone morphogenetic protein ( Bmp ) family [27] . Notch is essential for valve formation [28] , and the Notch pathway is activated by shear stress in HUVEC cells [29]–[31] . klf2a and bmp4 are expressed in the zebrafish conduction system [4] , [21] , [25] , . Our analysis concentrated on the AV canal during its specification ( between 22 and 48 hours postfertilization [hpf] ) [4] , [25] as well as slightly before valve leaflet formation ( 58 hpf ) [11] ( Figure 1G–1O ) . Both notch1b and klf2a were expressed in the endothelium ( Figure 1G–1L; Figure S2 , Video S3 ) . In contrast , bmp4 was expressed in the myocardium of the heart tube , starting around 20–22 hpf and later became restricted to the AV canal between 36 and 58 hpf ( Figure 1M–1O ) . Strikingly , expression of notch1b , klf2a , and bmp4 became restricted to the region of high reversing flow we identified in the AV canal as the heart matured . In the developing zebrafish heart , where the Reynolds numbers are much less than one [14] , flow patterns are dominated by the relationship between viscous forces and pressure gradients [32] . Thus , two methods of altering the reversing intracardiac blood flows in vivo are to: ( 1 ) manipulate blood viscosity , or ( 2 ) modulate pacemaker activity in order to change intracardiac pressure gradients [33] . To alter blood viscosity , we lowered the hematocrit by targeting two genes controlling early hematopoiesis in zebrafish , gata1 and gata2 [34] , with MOs . Embryos injected with gata1 MO are completely devoid of circulating blood cells [34] , have a lower blood viscosity ( reduced by ∼90% , see Materials and Methods ) , and display an increased RFF compared to controls ( Video S4; ∼RFF: 45%±12% in gata1 morphants , compared to 35%±7% in controls , Figure 2A and 2B ) . Embryos injected with gata2 MO contain fewer circulating blood cells in comparison to wild-type embryos ( 72% fewer blood cells , Figure 2L , Video S5 ) , have a reduced viscosity ( ∼70% lower than controls ) , and display a strongly reduced RFF ( 17%±4% of the heart cycle , Figure 2C; Video S4 ) in a majority of embryos ( n = 13 , 54% ) , highlighting the nonlinear relationship between heartbeat frequency and viscosity . When analyzed at 96 hpf , a majority of gata1 embryos had normal valves ( 77% of embryos had normal valves , n = 13; Figure 2F and 2L ) ; in contrast , the majority of gata2 morphants displayed severe valve defects ( 64% of the embryos displayed abnormal valves , n = 14; Figure 2G and 2L ) . To make sure that the abnormal valve development was related to the lower RFF and not to other functions of gata2 , we analyzed the effect of simultaneously inactivating gata1 and gata2 . This treatment further reduced blood viscosity , restored the RFF to 50% ( Figure 2D , Video S4 ) , and rescued valve formation ( 87% of embryos had normal valves , n = 8; Figure 2H and 2L ) . We also confirmed that lack of blood cells does not affect heart chamber patterning and vascular development ( Figure S3 ) . Because shear force depends directly upon viscosity , the reduced blood viscosity resulting from the gata1 or gata1/2 MOs reduces the magnitude of the shear forces throughout the cardiovascular system with respect to normal or gata2 morphants ( Figure S4 ) . Thus , the normal valve development of the gata1 and gata1/2 morphants , and the abnormal valve development in the gata2 morphants show that reversing flows , rather than magnitude of shear stress alone , are critical for valve leaflet formation ( Figure 2M ) . To better define the effects of RFF alteration in the gata2 morphants , we used quantitative reverse transcriptase PCR ( qRT-PCR ) to study a set of flow-responsive genes . We compared expression of bmp4 , klf2a , notch1b , neuregulin1 ( nrg1 ) , and endothelin1 ( edn1 ) in wild-type , gata1 , and gata1/2 morphant embryos ( Figure 2I–2K ) . Their expression levels in the gata1 morphants remained close to the control baseline ( Figure 2I ) , as did their levels in gata1/2 morphants , except for a slight decrease in bmp4 expression ( about 2-fold , Figure 2K ) . In gata2 morphants , two genes were significantly down-regulated: klf2a ( about 5-fold reduction ) and notch1b ( about 2 . 5-fold reduction ) ; edn1 and nrg1 display a mild reduction ( about 1 . 5-fold reduction , Figure 2J ) . Since wall shear stress ( WSS ) is a major stimulus for endothelial cell response in vitro , we explored whether it is also associated with the developmental changes we observe in vivo . Blood cell velocity measurements were used to estimate the WSS generated in the AV canal in control and altered flow conditions ( summarized in Figure S4 ) . In all gata morphants , the WSS is decreased due to the reduced blood viscosity . Interestingly , although gata2 and gata1 morphants display comparable amounts of WSS , they have opposite valve phenotypes . Thus , WSS magnitude cannot be the only determining factor for valvulogenesis . To explore this relationship further , we analyzed four heart contractility mutants ( cx36 . 7 , myh6 , ttna , and sih; Figure S5A–S5E ) , and find that they have widely varying RFFs ( Video S6 ) . Furthermore , the mutants exhibiting a decreased RFF demonstrate both reduced klf2a expression ( Figure S5F–S5H ) and increased valve dysgenesis ( Figure S5C , S5D , S5I , and S5J ) . Thus , results from animals with reduced blood viscosity and with reduced heart contractility suggest that , for normal development of valves , the reversing nature of the WSS is more important than its magnitude . We further explored the relationship between RFF and valve development by using lidocaine , a sodium channel blocker , to decrease heart rate [35] , as well as increased temperature to increase heart rate [33] . Lidocaine increases the time from ventricular contraction to the atrial contraction of the next heartbeat , thus lengthening the period between the onset of the E wave ( early diastolic filling due to ventricular suction ) and A wave ( ventricular filling due to atrial contraction ) . Slowing the heart rate by only 30% reduced the RFF by as much as 60% ( Figure 3A ) . Similarly , warming the animal by 2–4°C sped up the heart and reduced the RFF ( Video S7 ) . Because lidocaine is easily applied and rinsed out , we could decrease the RFF for defined periods to find the stages at which oscillatory flow is critical for valve development . Starting at 24 , 36 , or 48 h of development , we incubated fish in lidocaine for either 12 or 24 h , after which the fish were returned to normal medium ( Figure 3B ) . When scored at 96 hpf , valve leaflets were evident in all control embryos ( no lidocaine exposure; Figure 3C; Video S8 ) . In contrast , fish in which lidocaine reduced the RFF displayed a range of valve defects ( Figure 3B , blue bars and red bars ) . Similar defects were observed after reducing the RFF with elevated temperature ( Figure 3B , yellow bars ) . In the subtlest defect manifestations , valve leaflets did not form ( Figure 3D; Figure S3B; Video S8 ) . In more extreme cases , the heart retained an immature tubular shape ( 17% , n = 36; Figure S6 ) . The defects cannot be from a side effect of the lidocaine , as slight warming of the animals to restore heart rate , and thereby RFF , to normal rescued valve leaflet formation ( Figure 3E , 3H , and 3K , Video S8 ) . The highest proportion of valve defects resulted when 12- or 24-h lidocaine treatments were initiated at 36 hpf , suggesting the greatest sensitivity to decreased RFF from 36–48 hpf ( Figure 3B ) . Interestingly , this time window corresponds to the period when bmp4 , notch1b , and klf2a normally become restricted to the AV canal . To explore the timing relationships between the flow-responsive genes , we analyzed their expression by ISH after a 5- or 15-h lidocaine treatment , starting at 31 hpf . klf2a expression significantly decreased in as little as 5 h of treatment ( Figure 3 , compare 3F and 3G ) , indicating that klf2a may be an immediate target of the mechanism ( s ) that sense RFF . In contrast , expression of notch1b was normal after the short lidocaine treatment , but was decreased after 15 h of treatment ( Figure S7 ) . Quantitative PCR studies show that klf2a , edn1 , and notch1b were strongly down-regulated after 10 h of lidocaine treatment , started at 36 h , ( about 5-fold reduction compared to controls ) ; whereas nrg1 and bmp4 expression levels were almost normal ( Figure 3L and 3M ) . Shorter treatments ( 6 h ) led to a significant decrease in klf2a and nrg1 mRNA levels ( about 2 . 5- and 2-fold reduction , respectively ) , suggesting that these two genes may be primary targets of retrograde flow ( Figure 3L ) . The strong dependence of klf2a expression on the presence of oscillatory flow during both the 6- or 10-h treatments , as well as the similarity of its expression kinetics to those observed in cell culture [36] , makes klf2a an excellent candidate as a key component in mediating the effects of oscillatory flow on valve specification , validating the proposed involvement of this gene in vertebrate cardiogenesis [37]–[39] . We tested whether klf2a is required for valve formation by knocking down its expression using MOs , and obtained AV valve dysgenesis phenotypes that were remarkably similar to those of embryos exposed to reduced oscillatory shear stress ( scored at 96 hpf; Figure 4A–4D ) ; 52% of klf2a MO-treated embryos ( n = 36 ) revealed valve dysgenesis; none of the sham- or control-injected embryos ( n = 45 ) showed abnormal valve development ( Figure 4A–4D ) . This similarity in phenotypes suggests that expression of klf2a is a key part of the genetic program that makes valve development responsive to normal oscillatory flow ( Figure S8 ) . In mouse , loss of Klf2 is associated with heart failure and altered cardiac output [38] . In our studies , the zebrafish klf2a morphants displayed a heart rate similar to that of the control embryos at 48 hpf ( 1 . 7 Hz; Figure 4E and 4F ) , and had normal flow patterns within the AV canal ( n = 5 , Figure 4E and 4F ) . We found that atrial and ventricular fates are properly assigned in the klf2a morphants , because the chamber-specific expression of nppa , bmp4 , and cmlc2 appear normal ( Figure 4G and 4H; Figure S9A–S9D ) . Thus , the effects of our MO experiments are not secondary to an alteration in heart structure or blood flow . ISH revealed that the first apparent molecular defects in the klf2a morphants are a decrease in notch1b expression at 36 hpf and a lack of notch1b expression at the AV boundary of the heart at 46 hpf ( Figure 4L and Figure S9E and S9F ) , consistent with previous work showing that klf2 lies upstream of Notch in HUVEC cells [40] . In contrast , bmp4 expression is normal at 36 hpf and slightly decreased at 46 hpf ( Figure 4I and 4J and Figure S9C and S9D ) . When measured by qRT-PCR , the expression levels of bmp4 , edn1 , and nrg1 were lower than normal by at least a factor of two ( Figure 4M ) ; the strong decrease in notch1b expression seen by ISH corresponds to a 10-fold reduction compared to controls ( Figure 4M ) . Together with the fact that klf2a expression is a primary target of oscillatory flow , these data indicate that klf2a functions upstream of many known flow-induced genes in the process of AV valve formation in response to oscillatory flow . Zebrafish valves emerge from the endothelium through the combined actions of cell rearrangements and cell shape changes [1] , [11] . To characterize the leaflet phenotype in the different mutants exhibiting altered RFF , we analyzed cell number and cell shape using the Tg ( flk1:gfp ) fish line [41] at 72 hpf , a stage at which the valve invagination is clearly visible [11] . In this line , the GFP accumulates in endothelial cells , but the fluorescence level is different from cell to cell . The inherent brightness variation allows us to count and assess the shape of every endothelial cell in the heart . Focusing our analysis on the AV canal , we found that the endothelial ring forms in every morphant and lidocaine-treated embryo ( Figure 5A–5E ) . Strikingly , gata1 and gata1/2 morphants display normal cell numbers in the AV canal ( Figure 5A and 5B and unpublished data ) ; however , all treatments that disrupt the invagination of valve leaflets ( gata2 MO , klf2a MO , and lidocaine ) exhibit decreased endothelial cell number in the AV canal compared to the controls . Three-dimensional volumetric measurement of the endothelial AV cells reveals that wild-type controls as well as the gata1 and gata1/2 morphants possess endothelial cells that are cuboidal ( Figure 5A and 5B , 5F–5G , and 5Q; Video S9 ) ; in contrast , the endothelial cells remain flat and elongated in gata2 morphants , klf2a morphants , and lidocaine-treated embryos ( Figure 5C–5E , 5H–5J , and 5Q; Video S9 ) . These differences precede the absence of valve invagination in embryos with decreased RFF and suggest that cell remodeling is important for leaflet morphogenesis . Taken together , our results show that the loss of klf2a expression , lack of invagination , decreased endothelial cell number , and abnormal endothelial cell shape characterize the effects of decreased RFF .
In zebrafish , the first step of valvulogenesis involves the clustering of endothelial cells at the AV boundary . Cells coalesce to form an endothelial ring lining the AV canal between 24 and 48 hpf [6] . As seen previously [6] , we found that flow is not necessary for endocardial ring formation . However , blood flow is critical for cell shape change and leaflet invagination . Knocking down klf2a does not affects endothelial ring formation , confirming that klf2a function starts when its expression becomes detectable in the AV canal . Our data together with those of others [11] show that the endothelial ring is assembled in a region coinciding with klf2a expression , and that reversing flows progressively increase in amplitude specifically at the AV canal after the endothelial ring forms . This timing suggests that the effects of blood flow act after an initial patterning that is guided by a genetic program , reminiscent of the process acting in vascular development [46] . Thus , it appears that the earliest steps of heart development can be considered as genetically hardwired , but that secondary events , such as valvulogenesis , are contingent on the presence of reversing flows . Zebrafish valve development appears to be somewhat divergent from the process described in amniote vertebrates . In chicken and mice , valve leaflets arise from a mesenchymal cushion; in zebrafish , valves emerge directly through an invagination of the AV endothelium [11] . The origins of this morphogenetic process are unclear , but it allows the maturation of a functional valve in less than 96 h of embryonic development [11] , [22] . Our results show that this morphogenetic mechanism is dependent on reversing flow forces . Interestingly , the absence of invagination correlates with a lack of cell shape change that would normally occur during this process . Many observations using endothelial cell culture have shown that the presence of flow activates signaling pathways implicated in cytoskeletal remodeling [17] , [31] . It is thus tempting to speculate that reversing flows initiate the invagination process by stimulating the necessary movements and cytoskeletal rearrangements of endothelial cells in the AV canal to build a functional valve . The formation of heart valves allows unidirectional flow to be sustained as the peripheral vasculature develops and the increase in systemic resistance reduces the net flow of the valveless heart that results in the appearance of retrograde flow . The RFF is greater in the AV canal than in the rest of the heart or the cardiovascular system . The AV canal , a constriction , is exposed to high hemodynamic forces due to the higher velocities generated in areas with reduced cross section . Our studies clearly show that , although the drop in WSS magnitude affects gene expression levels in the heart , they are not sufficient to explain the abnormality in valve formation . Another aspect of the WSS , namely its oscillating directionality due to reversing flows , has to be included to understand the apparition of valve abnormalities in the gata2 morphants and not in the gata1 morphants where the shear forces are the lowest . Our quantitative imaging analysis strongly suggests that reversing flows are the proper stimulus controlling valve formation . Reversing flows have been observed in the developing cardiovascular system of many vertebrates ( [47] and S . E . Fraser , unpublished data ) and could be involved in other important steps of cardiovascular development . Among the many genes responsive to flow , klf2a seems to be specifically responsive to disturbed flows as observed both in vivo ( this study ) and in vitro [36] . Although , a direct involvement of klf2 ( the homolog of klf2a in mouse ) during valve development remains to be uncovered in higher vertebrates , this study should stimulate investigation of subtler valve alterations in these mutants [38] . Genetic evidence also suggests that klf2 has atheroprotective roles in adult mice [48] and humans [21] , further suggesting that klf2 , reversing flows , and cardiac physiology and development are tightly interconnected and that klf2 could also be implicated in the flow response during these processes . klf2a stands out as a possible early indicator of defective valve development . Nevertheless , it is clear that a number of other genes are mediating the response of endothelial cells to flow and that more investigations will be required to isolate them all as well as determine their interconnections . Given that heart-pumping activity and blood content constantly change as the heart develops , a patterning mechanism based on flow sensing provides a very practical way to coordinate the timing of valve formation with the pumping efficiency of the heart . In the context of valvulogenesis , reversing flows constitute an efficient signal by providing specific stimuli that dynamically locate the valve forming area . This hypothesis is fully consistent with emerging models arising from studies addressing the role of biomechanical stimuli during embryogenesis , which suggest that extrinsic forces and intrinsic hardwired programs are interconnected into feedback loops [49] , [50] . The advantage of such a mechanogenetic interplay is that it offers the opportunity for cells to locally adjust to the rapid environmental changes occurring in dynamic environments in conjunction with organizing centers [51] . In such systems , cells can directly react to the dynamics of the organ and can properly adapt at the single-cell level to organize as a coherently growing tissue . In conclusion , we have demonstrated that heart rate and blood viscosity can modulate the duration of oscillatory flow in vivo and have presented a set of useful methods to control hemodynamic forces during cardiogenesis . Together , these simple approaches offer powerful tools for predicting and potentially treating dysgenesis of cardiac valves and broaden the array of mechanisms to consider for explaining the origins of congenital cardiac malformation .
The Zeiss LSM 510 was used to image Tg ( flk1:gfp ) and BODIPY-ceramide ( Molecular Probes ) stained embryos to visualize valve structure . Embryos were anesthetized prior to imaging in 0 . 0175% tricaine and placed in agarose wells . All images were taken with a 40×/1 . 1 LD C-Apochromat water immersion lens . For high-speed imaging , the Zeiss LSM 5 LIVE was used to image BODIPY-ceramide–stained embryos and to visualize valve formation and flow patterns . Images of 256×256 pixels were acquired at 151 frames per second . Brightfield images were taken with a Basler A602f CMOS camera mounted on a home-built microscope equipped with an Olympus 0 . 5 NA 10× air objective coupled with a 300-mm focal length tube lens . Images were acquired at 216 frames per second . Transvalvular blood flow was characterized as positive , negative , or absent ( no flow ) by analyzing blood cell motions across the developing valve leaflets . For embryos lacking blood cells , the plasma was labeled by injecting microbeads ( Bangs Laboratories ) into the yolk sac . The region of interest was defined relative to the atrium and ventricle and moved with the valve plane during the cardiac cycle . Blood flow direction was marked in every frame taken during the cardiac cycle , and the retrograde flow fraction ( RFF ) was determined by dividing the total number of frames exhibiting retrograde flow by the total number of frames per cycle . For each treatment , five to 15 embryos were analyzed . The boxes represented in Figures 1 , 2 , and 4 represent the average flow observed during a minimum of ten heartbeats . Instantaneous blood cell velocity as a function of heart cycle time in the developing heart was assessed at 48 hpf by tracking blood cells manually in the AV canal over an average of four frames . Two heartbeats were analyzed in each condition . Shear stress was calculated as in [14] . The velocity of blood in the heart was modeled aswhere U is the centerline velocity , a is the half-width of the region of interest ( that is , the radius ) , and y is the distance from the wall . The shear stress iswhere μ is the dynamic viscosity of the fluid with units g·cm−1·s−1 . We measured the AV canal diameter every ten frames to set a ( on average a = 5 µm ) . The force exerted on a cell of surface area A is We assumed that the size of a cell in the AV canal was constant using 10 µm2 as its exposed surface area . The energy expenditure during one cardiac cycle ( E , in dyne·cm ) on a single cell is therefore given bywhere RFF is the retrograde flow fraction and f is the heart rate ( s−1 ) . Blood cells were imaged within the eye capillary . We counted the number of cells crossing a virtual line during the same time window in controls and MO-treated embryos ( Video S5 ) . Blood viscosity in gata morphants was estimated using the plot of relative viscosity versus particle volume fraction [52] after measurement of the particle volume fraction assuming fish blood composition is similar to that of humans . Heart rates of experimental embryos were decreased by dosage-dependent exposure to lidocaine added to the bathing solution . Lidocaine was drawn from the stock solution ( 1% stock , Abbott Laboratories ) and diluted into wells containing artificial pond water and approximately five embryos . Embryos were exposed to lidocaine for 24 h beginning at 31 hpf , the developmental stage marked by the transition from unidirectional to bidirectional flow . Assays of valve morphology and function were carried out at 96 hpf , a stage in which all wild-type fish hearts have at least one well-developed valve leaflet . Surviving embryos ( >80% ) were washed three times , placed in artificial pond water , and incubated at 28 . 5°C until being imaged ( 4 dpf ) . Normalized heart rates were calculated by dividing the heart rates of individuals ( n = 30 ) exposed to lidocaine by the heart rates of individuals under control conditions . Heart rates were measured after 1 h of continuous exposure to lidocaine ( Figure 2A ) . Zebrafish heart rates are regulated by ambient temperature . Unless otherwise noted , embryos were incubated at 28 . 5°C ( VWR Scientific incubator , model 2030 ) . To increase heart rates , a higher temperature ( 32 or 34°C ) incubator ( Thermolyne , model 37900 ) was used . Edema and abnormal cardiogenesis were observed when embryos were raised at 16°C , 20°C , and 35°C . Two MOs targeted against the putative translational site of klf2a were obtained from Gene Tools LLC ( 5′-gtaaaatcgttccactcaaagccat-3′-MO1; 5′-agctgagatgcatggacctgtccag-3′-MO2 ) . MOs were dissolved in 5 mM Hepes ( pH 7 . 6 ) and were injected into one-cell stage embryos ( total amount of 7 or 15 ng per embryo ) . We found that the two MOs induced the same range of malformations ( valve malformation: 40% , n = 15 for MO1; 52% , n = 36 for MO2; edema: 33% , n = 84 for MO1; 36% , n = 86 for MO2 ) . The specificity of each MO was assessed using a standard eGFP fusion approach where the eGFP sequence ( pEGFP-N1 , Clontech ) was fused by amplifying eGFP via PCR using primers containing the target sequence of each MO and a sp6 sequence in order to translate the PCR product ( mMESSAGE mMACHINE sp6 , Ambion ) ( Figure S10 ) . Control embryos were injected with a similar amount of a standard mismatch MO provided by Gene Tools LLC ( 5′-agGtgaCatgcatCgacctCtcgag-3′ ) . The specificity of this MO was addressed using the eGFP fusion approach ( Figure S11 ) , and its effect on valve development was analyzed using Tg ( flk1:EGFP ) embryos ( Figure S11 ) . Specificity of the MOs was further assessed by analyzing the ability of klf2a mRNA overexpression to rescue the MO-induced edema . A total of 100 pg of mRNA was coinjected with 15 ng of each MO , and edema was scored at 32 hpf ( Figure S10 ) . MOs to gata1 , gata2 and gata1/2 were used as in [34] , cx36 . 7 as in [15] , and myh6 as in [13] . ISHs were performed as described in [53] using the following probes: cmlc2 , bmp4 ( both provided by L . Trinh , California Institute of Technology ) , notch1b ( provided by M . Lardelli , University of Adelaide ) , nppa ( provided by T . Zhong , Vanderbilt Medical School ) , and klf2a probe ( obtained by PCR amplification of the plasmid IRBOp991B0734D provided by RPDZ , Berlin ) . A random sample of experimentally manipulated embryos was imaged at 96 hpf and scored based on the presence of valve leaflets . A focal plane with the atrium , ventricle , and AV canal in view was chosen to illustrate the phenotype . In cases where leaflets were difficult to identify ( <2% ) , the presence or absence of transvalvular retrograde flow was used to determine abnormal or normal valvulogenesis , respectively . A random sample of experimentally manipulated Tg ( flk1:EGFP ) embryos was imaged at 72 hpf , and a section plan of 10 µm was made using the substack maker plugin with Image J . Cell shapes were reconstructed in three dimensions using the contour surface key in Imaris ( Bitplane ) . A minimum of two embryos and ten cells in each condition were reconstructed . We then calculated the ratio of the length of the two longest sides and used a Z-test for two proportions to perform the statistical analysis . At 56 hpf , embryonic hearts were dissected in egg water after MO injection or lidocaine treatment . Two to three batches of ten hearts for each condition were pooled , and RNA was extracted using Trizol . RT was performed using the same amount of extracted mRNA and further tested by RT-PCR using the 96-well plate ABI 7000 QPCR machine in a SYBR Green ( Bio-Rad ) assay . The fold changes were calculated by the DCt method using a reference gene ( zebrafish TBP ) and plotted as a percentage of expression normalized to control . ANOVA tests were performed using Instat ( Graphpad Software , Inc ) . | The growth and development of vertebrates are critically dependent on efficient cardiac output to drive blood circulation . An essential step of heart development is the formation of heart valves , whose leaflets are made through a complex set of cellular rearrangements of endothelial cells . Endothelial cells experience high flow forces as blood circulates . Moreover , heart valves and associated structures can be malformed when flow forces are abnormal , suggesting that these flow forces are in fact required for proper valve formation . Whether it is the force of the blood flow , its directionality ( forward or reverse ) , or both that are important is not clear . We studied the interplay during valve development between key genes known to be involved in the process and epigenetic influences such as flow forces . Using zebrafish , whose optical clarity allows analyzing blood flow patterns at high resolution , we identified the presence of reversing flows specifically at the level of valve precursors . By manipulating blood flow patterns , we show that reversing flows are essential for valve morphogenesis . Specifically , we show that the expression of the gene klf2a depends on the presence of reversing flows and is required for valve development . We predict that by influencing levels of klf2a , reversing flows constitute an important stimulus controlling the appropriate biological responses of endothelial cells during valve formation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cardiovascular",
"disorders/hemodynamics",
"developmental",
"biology/morphogenesis",
"and",
"cell",
"biology"
] | 2009 | Reversing Blood Flows Act through klf2a to Ensure Normal Valvulogenesis in the Developing Heart |
Despite recent interest in reconstructing neuronal networks , complete wiring diagrams on the level of individual synapses remain scarce and the insights into function they can provide remain unclear . Even for Caenorhabditis elegans , whose neuronal network is relatively small and stereotypical from animal to animal , published wiring diagrams are neither accurate nor complete and self-consistent . Using materials from White et al . and new electron micrographs we assemble whole , self-consistent gap junction and chemical synapse networks of hermaphrodite C . elegans . We propose a method to visualize the wiring diagram , which reflects network signal flow . We calculate statistical and topological properties of the network , such as degree distributions , synaptic multiplicities , and small-world properties , that help in understanding network signal propagation . We identify neurons that may play central roles in information processing , and network motifs that could serve as functional modules of the network . We explore propagation of neuronal activity in response to sensory or artificial stimulation using linear systems theory and find several activity patterns that could serve as substrates of previously described behaviors . Finally , we analyze the interaction between the gap junction and the chemical synapse networks . Since several statistical properties of the C . elegans network , such as multiplicity and motif distributions are similar to those found in mammalian neocortex , they likely point to general principles of neuronal networks . The wiring diagram reported here can help in understanding the mechanistic basis of behavior by generating predictions about future experiments involving genetic perturbations , laser ablations , or monitoring propagation of neuronal activity in response to stimulation .
Determining and examining base sequences in genomes [1] , [2] has revolutionized molecular biology . Similarly , decoding and analyzing connectivity patterns among neurons in nervous systems , the aim of the emerging field of connectomics [3]–[6] , may make a major impact on neurobiology . Knowledge of connectivity wiring diagrams alone may not be sufficient to understand the function of nervous systems , but it is likely necessary . Yet because of the scarcity of reconstructed connectomes , their significance remains uncertain . The neuronal network of the nematode Caenorhabditis elegans is a logical model system for advancing the connectomics program . It is sufficiently small that it can be reconstructed and analyzed as a whole . The neurons in the hermaphrodite worm are identifiable and consistent across individuals [7] . Moreover the connections between neurons , consisting of chemical synapses and gap junctions , are stereotypical from animal to animal with more than reproducibility [7]–[10] . Despite a century of investigation [11] , [12] , knowledge of nematode neuronal networks is incomplete . The basic structure of the C . elegans nervous system had been reconstructed using electron micrographs [7] , but a major gap in the connectivity of ventral cord neurons remained . Previous attempts to assemble the whole wiring diagram made unjustified assumptions that several reconstructed neurons were representative of others [13] . Much previous work analyzed the properties of the neuronal network ( see e . g . [14]–[20] and references therein and thereto ) based on these incomplete or inconsistent wiring diagrams [7] , [13] . In this paper , we advance the experimental phase of the connectomics program [6] , [21] by reporting a near-complete wiring diagram of C . elegans based on original data from White et al . [7] but also including new serial section electron microscopy reconstructions and updates . Although this new wiring diagram has not been published definitively before now , it has already been freely shared with the community through the WormAtlas [22] and has also been used in previous studies such as [23] . See Methods section for further details on the wiring diagram and on freely obtaining it in electronic form . We advance the theoretical phase of connectomics [24] , [25] , by characterizing signal propagation through the reported neuronal network and its relation to behavior . We compute for the first time , local properties that may play a computational purpose , such as the distribution of multiplicity and the number of terminals , as well as global network properties associated with the speed of signal propagation . Unlike the conventional “hypothesis-driven” mode of biological research , our work is primarily “hypothesis-generating” in the tradition of systems biology . Our results should help investigate the function of the C . elegans neuronal network in several ways . A full wiring diagram , especially when conveniently visualized using a method proposed here , helps in designing maximally informative optical ablation [26] or genetic inactivation [27] experiments . Our eigenspectrum analysis characterizes the dynamics of neuronal activity in the network , which should help predict and interpret the results of experiments using sensory and artificial stimulation and imaging of neuronal activity . Organization of the Results section reflects the duality of contribution and follows the tradition laid down by genome sequencing [1] , [2] . We start by describing and visualizing the wiring diagram . Next , we analyze the non-directional gap junction network and the directional chemical synapse network separately . We perform these analyses separately because understanding the parts before the whole provides didactic benefits and because this delays making assumptions about the relative weight of gap junctions and chemical synapses . Finally , we analyze the combined network of gap junctions and chemical synapses .
For quantitative characterization , we first consider the gap junction network . Now we consider the chemical synapse network . Recall that due to structural differences between presynaptic and postsynaptic ends of a chemical synapse , electron micrographs can be used to determine the directionality of connections . Hence the adjacency matrix is not symmetric as it was for the gap junction network . Having considered the gap junction network and the chemical synapse network separately , we also examine the two networks collectively . To study the two networks , one may either look at a single network that takes the union of the connections of the two networks or one may look at the interaction between the two networks . Although the reported wiring diagram corrects errors in previous work and is considered self-consistent , one might wonder how remaining ambiguities and errors in the wiring diagram might affect the quantitative results presented . Furthermore there are connectivity pattern differences among individual worms; these individual variations may have similar effects on the analysis as errors and ambiguities . For network properties that are defined locally , such as degree , multiplicity , and subnetwork distributions , clearly small errors in the measured wiring diagram lead to small errors in the calculated properties . For global properties such as characteristic path length and eigenmodes , things are less clear . To study the robustness of global network properties to errors in the wiring diagram , we recalculate these properties in the wiring diagrams with simulated errors . We simulate errors by removing randomly chosen synaptic contacts with a certain probability and assigning them to a randomly chosen pair of neurons . Then , we calculate the global network properties on the ensemble of edited wiring diagrams . The variation of the properties in the ensemble gives us an idea of robustness . First , we explore the robustness of the small world properties and the giant component calculations . We edit wiring diagrams by moving each gap junction contact with probability and chemical synapse contact with probability . Tables 5 and 6 in Text S4 show the global properties for random networks obtained by editing the experimentally measured network . These tables suggest that our quantitative results are reasonably robust to ambiguities and errors in the wiring diagram . Properties for the neuronal network from prior work in [13] are also shown for comparison . The number of synaptic contacts that must be moved to achieve this network ( editing distance ) roughly corresponds to that with probability . Second , we characterize robustness for the linear systems analysis . Because of greater sensitivity of the eigenvalues to errors , we edit wiring diagrams by moving each gap junction contact with probability and a chemical synapse contact with probability . The spectra for randomly edited networks along with the spectrum for the measured network ( Figure 8 ( a ) ) are shown in Figure 10 . Although the locations of eigenvalues shift in the complex plane , many of them move less than the nearest neighbor distance and remain isolated . In addition to considering the effect of typical random edits , we can characterize the effect of worst-case errors on the eigenvalues using the -pseudospectrum [75] , which gives the eigenvalue loci for all perturbations by matrices of norm ( Figure 10 ) . For the gap junction , is simply the set of disks of radius around the eigenvalues , but for the chemical and combined networks , and are larger . In the worst case scenario , most eigenmodes become mixed up . Electron micrographs of chemical synapses have a further ambiguity when more than one postsynaptic partner receives input at a release site . We treated such polyadic ( send_joint ) synapses no differently than other synapses , but one might alternatively determine multiplicity by counting such synapses at strength . This alternate quantitation clearly does not change statistics that ignore multiplicity; the change in the spectrum is depicted in Figure 10 . Small deviations from equality when weighting gap junctions and chemical synapses to form the combined network yield similar spectral changes as the alternate quantitation of chemical synapses displayed in Figure 10 .
We have presented a corrected and more comprehensive version of the neuronal wiring diagram of hermaphrodite C . elegans using materials from White et al . [7] and new electron micrographs . Despite the significant additions , this wiring diagram is still incomplete due to methodological limitations discussed in the An Updated Wiring Diagram section . Yet , our work represents the most comprehensive mapping of the neuronal wiring diagram to date . The sensitivity of our analysis to methodological limitations ( and to network structure variation among individual organisms ) is discussed in the Robustness Analysis section . We proposed a convenient way to visualize the neuronal wiring diagram . The corrected wiring diagram and its visualization should help in planning experiments , such as neuron ablation . Next , we performed several statistical analyses of the corrected wiring , which should help with inferring function from structure . By using several different centrality indices , we found central neurons , which may play a special role in information processing . In particular , command interneurons responsible for worm locomotion have high degree centrality in both chemical and gap junction networks . Interestingly , command interneurons are also central according to in-closeness , implying that they are in a good position to integrate signals . However , most command interneurons do not have highest out-closeness , meaning that other out-closeness central neurons , such as DVA , ADEL/R , PVPR , etc . , are in a good position to deliver signals to the rest of the network . Linear systems analysis yielded a principled methodology to hypothesize functional circuits and to predict the outcome of both sensory and artificial stimulation experiments . We have identified several modes that map onto previously identified behaviors . Networks with similar statistical structural properties may share functional properties thus providing insight into the function of the C . elegans nervous system . To enable comparison of the C . elegans network with other natural and technological networks [76] , we computed several structural properties of the neuronal network . In particular , the gap junction network , the chemical synapse network , and the combined neuronal network may all be classified as small world networks because they simultaneously have small average path lengths and large clustering coefficients [14] . The tails of the degree and terminal number distributions for the gap , chemical and combined networks ( with the exception of the in-numbers ) follow a power law consistent with the network being scale-free in the sense of Barabási and Albert [40] . The tails of some distributions can also be fit by an exponential decay , consistent with a previous report [15] . However , we found that exponential fits for the tails have ( sometimes insignificantly ) lower log-likelihoods than power laws , making the exponential decay a less likely alternative . For whole distributions , neither distribution passes the -value test; if one is forced to choose , the exponential decay may be a less poor alternative . Several statistical properties of the C . elegans network are similar to those of the mammalian cortex . In particular , the whole distribution of C . elegans chemical synapse multiplicity is well-fit by a stretched exponential ( or Weibull ) distribution ( Figure 6 ( d ) ) . Taking multiplicity as a proxy of synaptic connection strength , this is reminiscent of the synaptic strength distribution in mammalian cortex , which was measured electrophysiologically , [30] , [77] . The definition of stretched exponential distribution is slightly different [30] , but has the same tail behavior . The stretch factor is , close to that in the cortical network . In addition , we found that motif frequencies in the chemical synapse network are similar to those in the mammalian cortex [77] . Both reciprocally connected neuron pairs and triplets with a connection between every pair of neurons ( regardless of direction ) are over-represented . The similarity of the connection strength and the motif distributions may reflect similar constraints in the two networks . Since proximity is unlikely to be the limiting factor , we suggest that these constraints may reflect functionality . We found that the chemical synapse and the gap junction networks are correlated , which may provide insight into their relative roles . To conclude the paper , let us note that our scientific development was not hypothesis-driven , but rather exploratory . Yet we hope that the reported statistics will help in formulating a theory that explains how function arises from structure .
This section describes the methods used to determine neuronal connectivity; see [78] for further details . We started assembling the wiring diagram by consolidating existing data from both published and unpublished sources . Using J . G . White et al . 's The Mind of a Worm ( MOW ) [7] as the starting point , we extracted wiring data from diagrams , figures , tables , and text ( for example , see [7 , Appendix A , pp . 118–122] on neuron AVAL/R ) . The connectivity of each neuron , its synaptic partner , and synaptic type ( chemical , gap junction , neuromuscular ) was manually entered into an electronic database . In the ventral cord , determining this level of synaptic specification was complicated by the fact that connections were recorded by neuron class . For example , bilateral neurons PVCL and PVCR were simply listed as PVC . We assigned proper connections to the appropriate left/right neuron by referring to White and coworker's original laboratory notebooks and original electron micrographs . In some cases , the number of synapses for a given neuron class in MOW differed from the sum of connections for the bilateral pairs in the notebooks and/or electron micrographs . The synaptic value of these neurons was determined by taking the value in MOW and dividing it between the left/right neurons proportionally to the values in the notebooks and/or electron micrographs . Next we incorporated R . M . Durbin's data for the anterior portion of the worm , reconstructed from animal N2U [8] . For neurons that projected beyond the nerve ring , only the anterior connections needed update . Since data from MOW did not specify the location of synapses , integration proved difficult . For these neurons , we obtained positional information by cross-referencing Durbin's data against original electron micrographs and his handwritten annotations in White's laboratory notebooks . Only synapses located in regions addressed by Durbin were included . Connections in the middle and tail regions of the worm were mostly unaffected by these updates . Studies based on green fluorescent protein ( GFP ) reporters mostly confirm the electron micrograph reconstructions described in MOW . A few differences between GFP-stained neurons and White's work have been observed [Hobert O and Hall DH , unpublished] . Notably , the anterior processes of DVB and PVT could have been mistakenly switched in MOW [7] . Based on these findings , we reversed the connections for neurons DVB and PVT anterior to the vulva . Most published works have focused on the neck and tail regions of C . elegans where most neuron cell bodies reside . Reconstructions of neurons in the mid-body of the worm , on the other hand , are scant and incomplete . From a combination of published works [7] , [8] , [10] , [79] , we found that wiring data for neurons had large gaps or were missing entirely . Sixty-one of these were motor neurons in the ventral cord . Two were excretory neurons ( CANL/R ) that do not appear to make any synapses . The remaining neuron , RID , is the only process in the dorsal cord that extends over the length of the animal . At the C . elegans archive ( Albert Einstein College of Medicine ) , we uncovered a large number of reconstruction records in White et al . 's laboratory notebooks . These notebooks identified neurons by different color code labels depending on the animal , the location of the neurite ( ventral or dorsal ) , and magnification of the electron micrograph . To ascertain the identity of the neurons , we relied on a combination of color code tables and comparisons of common anatomical structures between electron micrograph prints . In the end , we identified notes for full reconstructions of of the aforementioned neurons . Partial connectivity data for the remaining were also available where neurons have partial/missing dorsal side connections and neurons have partial ventral side connections . We checked the connections of all ( both published and unpublished ) neurons in the ventral cord against electron micrographs used by White and coworkers . Over updates were made to the original notes and published reconstructions . Many of these updates were additions of previously missed neuromuscular junctions between ventral cord motor neurons and body wall muscles . We found that a large section on the dorsal side of the worm , from just anterior to the vulva to the pre-anal ganglion , was never electron micrographed at high power magnification . This dearth of imagery was why so many neurons were missing dorsal side reconstructions . Using original thin sections for the N2U worm prepared by White et al . , we produced new high power electron micrographs of this dorsal region . Due to the condition of the sections , only one of every – sections was imaged . These new electron micrographs extended nearly on the dorsal side . New dorsal side data for 3 neurons ( DA5 , DB4 , DD3 ) were obtained from these electron micrographs . Resource constraints prevented us from covering the entire dorsal gap . From our compilation of wiring data , including new reconstructions of ventral cord motor neurons , we applied self-consistency criteria to isolate neurons with mismatched reciprocal records . The discrepancies were reconciled by checking against electron micrographs and the laboratory notebooks of White et al . Connections in the posterior region of the animal were also cross-referenced with reconstructions published in [10] . Reconciliation involved synapses for neurons ( chemical “sends , ” chemical “receives , ” and electrical junctions ) . For a random network with neurons and probability of a connection being present , if the constant , then the size of the giant component is asymptotically normal with mean and variance [80 , p . 120] . These quantities are given by ( 14 ) where ( 15 ) and is the Lambert -function . If we take to be and to be , then . Using the asymptotic approximation , the size of the giant component is distributed approximately normally with mean and variance . Thus the probability of having a giant component of size , which is over standard deviations from the mean , is about . If a precise evaluation of this infinitesimal value is desired , large deviations techniques , rather than the asymptotic approximation may be more valid [81] . To apply this method to the weakly connected component of a directed network , we are interested in the undirected network formed by adding a connection between two neurons if there is a connection in either direction . For a random directed network with probability of presence of a directed connection , the probability of a connection existing in either direction is . Taking to be , is . Then for an undirected random network with and the specified , is . Then the size of the giant component is distributed approximately normally with mean and variance . Thus the probability of the giant weakly connected component containing all the neurons in such a random network is overwhelming . Again , large deviations techniques should be used to get a precise evaluation of the probability of being on the order of standard deviations away from the mean . Consider the ensemble of random networks with a given degree distribution [82] . For the gap junction network , the generating function corresponding to the measured degree distribution iswith derivativeTherefore . The generating function is thenAs shown in [82] , the expected fraction of the network taken up by the giant component , , is , where is the smallest non-negative solution to . In our case , we find , and so . That is to say , one would expect the giant component to consist of neurons . Using the computed and , we can find the average component size excluding the giant component , which turns out to be . For the symmetrized chemical network , the generating function corresponding to the measured degree distribution iswith derivativeTherefore . The generating function is then The expected fraction of the network taken up by the giant component , , is , where is the smallest non-negative solution to . Here is found to be , and so . That is to say , one would expect the giant component to consist of neurons . Continuing from the previous subsection , we find the derivative of the generating function for the gap junction network to beThus . Letting and , it is shown in [82 , ( 53 ) ] , that the expected path length is ( 16 ) To find functional forms of the tails of various distributions , we follow the procedure outlined in [42] . For the candidate functional forms—say , the power law and the exponential decay —we perform the following steps . First , we find the optimal parameter of the fit by maximizing the log-likelihood and the optimal starting point of the fit by minimizing the Kolmogorov-Smirnov statistic . Second , we evaluate the goodness of fit by calculating the -value that the observed data was generated by the optimized distribution using as a criterion for plausibility . Finally , if several distributions pass the -value test we compare their log-likelihoods to find the most probable one . Let us bound the probability of finding an eigenmode that comprises a random set of neurons . Let be the number of neurons in the network being analyzed . Let be the number of neurons that appear strongly in the th eigenmode and let . Furthermore let be the number of neurons in the random set , which one might endeavor to investigate as a putative functional circuit derived from an eigenmode . Now go through each eigenmode and add to a list all possible unordered -tuples of strong neurons . Even if all of these are unique , the size of the list is upper-bounded by which itself is upper-bounded by . Additionally , we can compute the number of all unordered -tuples of neurons . This number is . Thus , if a random set of neurons was selected from all possible sets of neurons , the probability that there would be an eigenmode containing all of them is upper-bounded as Suppose we are interested in putative functional circuits of size in the giant component of the gap junction network , which has and from Figure 2 in Text S4 take . Then even the loosest upper-bound yieldsand so finding a random set of neurons in an eigenmode is unlikely . Suppose we know functional circuits of size through molecular biology and want to know the probability of at least one of them appearing in the eigenmodes by chance . By the union bound ( Boole's inequality ) , this probability is less than . If we take and , the probability of a known functional circuit appearing in the eigenmodes by chance is less than for the giant component of the gap junction network . The likelihood ratios shown in Figure 9 are the following quantities , empirically estimated from either all neuron pairs or pairs with a GABAergic presynaptic neuron . The first isThe second isand the third is Table 1 shows the Pearson correlation coefficients between neuron degree sequences . The average Pearson correlation coefficients of randomly permuted degree sequences from trials are also shown for comparison . The standard deviation is also shown since the distributions of the three randomized correlation coefficients were all nearly symmetric about zero . We used the MATLAB package EigTool [83] to compute pseudospectra . Note that MATLAB code for computing several network properties is available at http://mit . edu/lrv/www/elegans/ . This collection of software may be used not only to reproduce most of the figures in this paper , but also for future connectomics analyses . The collected data is available from the WormAtlas [22] as well as from the same website as the MATLAB code . | Connectomics , the generation and analysis of neuronal connectivity data , stands to revolutionize neurobiology just as genomics has revolutionized molecular biology . Indeed , since neuronal networks are the physical substrates upon which neural functions are carried out , their structural properties are intertwined with the organization and logic of function . In this paper , we report a near-complete wiring diagram of the nematode Caenorhabditis elegans and present several analyses of its properties , finding many nonrandom features . We give novel visualizations and compute network statistics to enhance understanding of the reported data . We also use principled systems-theoretic methods to generate hypotheses on how biological function may arise from the reported neuronal network structure . The wiring diagram reported here can further be used to generate predictions about signal propagation in future perturbation , ablation , or artificial stimulation experiments . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience/motor",
"systems",
"neuroscience/theoretical",
"neuroscience",
"neuroscience/sensory",
"systems"
] | 2011 | Structural Properties of the Caenorhabditis elegans Neuronal Network |
Close interpersonal contact likely drives spatial clustering of cases of cholera and diarrhea , but spatial clustering of risk factors may also drive this pattern . Few studies have focused specifically on how exposures for disease cluster at small spatial scales . Improving our understanding of the micro-scale clustering of risk factors for cholera may help to target interventions and power studies with cluster designs . We selected sets of spatially matched households ( matched-sets ) near cholera case households between April and October 2013 in a cholera endemic urban neighborhood of Tongi Township in Bangladesh . We collected data on exposures to suspected cholera risk factors at the household and individual level . We used intra-class correlation coefficients ( ICCs ) to characterize clustering of exposures within matched-sets and households , and assessed if clustering depended on the geographical extent of the matched-sets . Clustering over larger spatial scales was explored by assessing the relationship between matched-sets . We also explored whether different exposures tended to appear together in individuals , households , and matched-sets . Household level exposures , including: drinking municipal supplied water ( ICC = 0 . 97 , 95%CI = 0 . 96 , 0 . 98 ) , type of latrine ( ICC = 0 . 88 , 95%CI = 0 . 71 , 1 . 00 ) , and intermittent access to drinking water ( ICC = 0 . 96 , 95%CI = 0 . 87 , 1 . 00 ) exhibited strong clustering within matched-sets . As the geographic extent of matched-sets increased , the concordance of exposures within matched-sets decreased . Concordance between matched-sets of exposures related to water supply was elevated at distances of up to approximately 400 meters . Household level hygiene practices were correlated with infrastructure shown to increase cholera risk . Co-occurrence of different individual level exposures appeared to mostly reflect the differing domestic roles of study participants . Strong spatial clustering of exposures at a small spatial scale in a cholera endemic population suggests a possible role for highly targeted interventions . Studies with cluster designs in areas with strong spatial clustering of exposures should increase sample size to account for the correlation of these exposures .
Cholera is responsible for over 100 , 000 deaths each year and is endemic in many countries [1] . In Bangladesh , approximately 450 , 000 cases of cholera are estimated each year [2] . Cholera cases have been known to be tightly geographically clustered . For example , the scale of clustering of cholera cases in Matlab , Bangladesh was found to be around 250m in one study , and under 1km in another [3 , 4] . Close interpersonal contact likely drives clustering of secondary transmitted cases [5 , 6] . However , few studies have explored whether clustering in the risk factors themselves may explain clustering in disease . Understanding if and how risk factors cluster may help us better understand how cholera and diarrheal diseases cluster . Furthermore , cholera prevention strategies often consist of behavioral interventions aimed at encouraging communities to adopt safe water , sanitation and hygiene practices [7] . The presence of micro-scale clustering would suggest that intensive , highly targeted , hygiene interventions might be effective complements to more general campaigns . Taking into account clustering of risk factors in modeling the effect of vaccine on cholera transmission can help identify optimal populations for vaccine deployment and thus increase effectiveness of immunization campaigns . More importantly , clustering of risk factors has important implications for the design of both spatially matched case-control studies and clustered survey design . Clustering of risk factors at a fine resolution may lead us to match on factors other than those we intended to match on when selecting spatially matched controls , biasing risk estimates towards the null and reducing the power of the study to detect associations [8] . It would also require cluster survey designs to have an increased sample size in order to account for correlation in risk , thus having profound implications for cholera vaccine studies adopting this design . Here we use data from spatially matched households enrolled as controls during a case-control study to explore whether potential risk factors for cholera and other diarrheal diseases cluster at small spatial scales . The risk factors considered include lack of access to water , poor sanitation , overcrowding , source of drinking water , and poor hygiene and food handling practices [9–17] . These specific risk factors may or may not be responsible for cholera infection in the Arichpur neighborhood , but have been previously identified as important risk factors in Bangladesh and other countries [9–17] . We examine both the geographic clustering of individual risk factors , and whether groups of risk factors tend to appear together in individuals , households and nearby neighbors .
The present study used data from households selected as spatially matched controls in a case-control study conducted between April and October 2013 in Arichpur in Tongi Township in Dhaka , Bangladesh . Arichpur is a cholera endemic working class neighborhood , and is approximately 1 . 2 km2 with a population density over 100 , 000 per km2 [18] . Cholera cases were consenting Arichpur residents aged 2 years or older who presented to local pharmacies or hospitals with acute diarrhea and had V . Cholerae O1 or O139 cultured from their stool sample . Hospital surveillance was conducted at the Tongi Hospital and the icddr , b Dhaka Hospital , both frequented by Arichpur residents . Study staff attempted to enroll four spatially-matched control households ( the matched-set ) for each primary case household . To select a control household , staff would first select a number ( one for each control household ) from a random number table , then attempting to enroll the household that number of doors to the right of the entrance of the primary case household . Special instructions were applied for households in multi-story dwellings and when a natural boundary ( e . g . railroad or river ) was encountered . Household GPS locations were collected at the front door of each household ( resolution ~3-5m ) . During one of the two household visits to each control household , we collected data on exposure to suspected cholera risk factors at both the household and individual level . Demographic variables including household size and number of rooms were collected . We also collected data on hygienic behavior including the availability of hand soap as observed by the team , and whether the household members boiled all of their ( self-reported ) sources of drinking water . Exposures related to sanitation and latrine included type of latrine and number of households sharing a latrine . For access to drinking water , we measured: 1 ) whether the household consumed drinking water from municipal supplied water ( supplied water ) and/or tubewell water in the past month by self-report; 2 ) distance to water source from front door of the house as measured by the team; 3 ) whether households had intermittent access to drinking water; and 4 ) whether and how households stored drinking water ( See S1 Text for related questions from study questionnaire ) . Individual-level exposures were collected at each visit from each consenting household member age 2 years of older . The exposures included: feeding a child by hand , eating meals prepared over 2 hours before consumption , drinking water outside the home , eating fresh cut fruit or vegetables outside the home , and drinking tea outside the home . The original questionnaire asked for frequency of each exposure in the past week ( no exposure , 1–2 days , 3 or more days but not every day , or every day ) ( See S1 Text for related questions from study questionnaire ) . We dichotomized all exposures into high risk and low risk categories based on published literature on risk factors for cholera . Household density was defined as the ratio of number of household members to number of rooms , and was dichotomized at the mean . For other household level exposures , we determined that high risk categories were: using pit latrine ( vs . modern/septic tank/sanitary ) , sharing a latrine with other households , storing drinking water , a distance of > 10 meters to nearest drinking water source from front door , intermittent drinking water supply , not always boiling drinking water from all reported source , no hand soap available , drinking municipal supplied water , and drinking tubewell water [9–17] . Although drinking municipal supplied water is a marker of improved water source under the Millennium Development Goals , we characterized it as a risk factor because the “improved sources” would still carry risks if treated inadequately [19–21] . Each individual level exposure was dichotomized into the exposed vs . not exposed categories . Exposures that showed little variability between households and individuals ( or appeared in over 98% of households or individuals ) were excluded from the main analyses ( See S1 Text for details ) . We explored the clustering of each household level exposure within matched-sets by calculating the intraclass correlation coefficient ( ICC ) . For individual level exposures , we explored clustering within matched-sets as well as within individual households . We estimated the ICC using a two-level random effects logistic regression model with no additional covariates specified using the GLLAMM program in Stata 2012 [22 , 23] . We assumed a binomial distribution and specified logit link function for the binary exposures , representing the proportion of total variance in the underlying continuous latent response to each binary exposure that was due to the differences between matched-sets or households . We considered that the exposures with ICC estimates over 0 . 7 were strongly correlated , and the exposures with ICC estimates between 0 . 5 and 0 . 7 were moderately correlated . To understand whether clustering of each household and individual level exposure exhibited fine scale spatial dependence , we explored the association between the level of clustering in matched-sets and their geographic extent . We quantified the clustering within each matched-set ( within matched-set concordance ) as the proportion of pairs of households ( or individuals ) within the matched-set that had the same exposure . We defined the spatial extent of matched-sets as the median distance between any two households within the matched-set . The confidence interval of the association was assessed using bootstrapping ( 1000 iterations ) by resampling all matched-sets , households within the matched-sets , and then individuals within the households . We also explored whether clustering existed at larger spatial scales ( i . e . , distances greater than the spatial extent of matched-sets ) . We quantified concordance of exposures between two matched-sets ( between matched-sets concordance ) as the proportion of all possible pairs of households , one from each matched-set , that had the same exposure . We used non-parametric locally weighted polynomial regression models ( LOESS ) to visually assess how between matched-sets concordance for each exposure varied with the distance between centroids of matched-sets . We considered distances up to 780 meters ( approximately ¾ of the diameter of Arichpur ) as households and individuals that were separated by over 780 meters were limited . We hypothesized decreasing concordance between matched-sets for exposures that showed clustering beyond the spatial extent of the matched-sets . If concordance remained static beyond a certain distance threshold , the threshold likely corresponded to the spatial extent of the clustering . We used linear regressions to quantify the trend of decreasing concordance up to the threshold and calculated the confidence intervals using bootstrapping ( 1000 iterations ) by resampling all matched-sets , households within the matched-sets , and then individuals within the households . To understand which household level exposures clustered with one another , we calculated the probability of co-occurrence of different household level exposures within the same household , or matched-set . We then calculated the ratio of this probability to the probability of exposures appearing together by chance if they were uncorrelated . Likewise , we calculated the co-occurrence of each pair of individual level exposures within individuals , households and matched-sets ( See S2 Text for detailed description of methods ) . Values above one suggested two exposures tend to appear in the same individuals , households , or matched-sets closer to each other ( co-occurrence of two exposures ) , while values less than one suggested two exposures tend to not appear together . Confidence intervals were calculated by resampling all matched-sets , households within the matched-sets , and then individuals within the households through 1000 bootstrap iterations . We obtained written informed consent from adult study participants or parents/guardian for minors . The study protocol was reviewed and approved by the ethical review committee of Johns Hopkins Bloomberg School of Public Health and icddr , b .
We observed pronounced clustering of household level exposures within matched-sets ( Fig 2A ) . Neighboring households nearly always used the same type of drinking water ( ICC for use of supplied water = 0 . 97 , 95%CI = 0 . 96 , 0 . 98; ICC for use of tubewell water = 0 . 94 , 95%CI = 0 . 87 , 1 . 00 ) . Neighboring households frequently used the same type of latrine ( ICC for modern vs . pit latrine = 0 . 88 , 95%CI = 0 . 71 , 1 . 00 ) and had similar access to drinking water ( ICC intermittent water supply = 0 . 96 , 95%CI = 0 . 87 , 1 . 00 , and distance to water source 0 . 80 , 95%CI = 0 . 62 , 0 . 98 ) . Sharing a latrine ( ICC = 0 . 70 , 95%CI = 0 . 43 , 0 . 97 ) and hygienic practices ( ICC for soap availability = 0 . 68 , 95%CI = 0 . 48 , 0 . 88 ) showed moderate clustering . For individual behaviors , we only observed strong clustering of eating meals prepared over 2 hours before consumption ( ICC = 0 . 90 , 95% CI = 0 . 79 , 1 . 00 within households; ICC = 0 . 74 , 95% CI = 0 . 55 , 0 . 93 for within matched-sets ) . More spatially compact matched-sets tended to have higher concordance in exposures than the larger ones , though the trend was not statistically significant for any exposure ( Fig 2C and 2D ) . Averaging the concordance of all exposures in each matched-set , we observed that , for each 100 meter increase in the geographic extent of a matched-set , the within matched-sets concordance decreased by 0 . 13 ( 95%CI = 0 . 064 , 0 . 41 ) for household level exposures and by 0 . 069 ( 95%CI = -0 . 022 , 0 . 19 ) for individual level exposures . Visually assessing the LOESS plots of the exposures , we found clustering beyond the spatial extent of the matched-sets in household level exposures related to drinking water sources , as manifested by the decreasing concordance between matched-sets over space . The spatial extent of the clustering was approximately 350 meters for the use of supplied water and tubewell water , and 460 meters for distance over 10 meters to the closest water source ( Fig 3 ) , since concordance remained static beyond these thresholds . We also found an overall decreasing trend in concordance for intermittent water supply and boiling water . Considering distance up to the corresponding spatial extent of decreasing concordance , we observed that for each 100 meters increase in distances between two matched-sets , the concordance between matched-sets decreased by 0 . 11 ( 95%CI = 0 . 019 , 0 . 20 ) for the use of supplied water , and 0 . 097 ( 95%CI = 0 . 026 , 0 . 12 ) for the distance to the closest drinking water source . We did not observe strong decreasing concordance with distance in individual level exposures ( S1 Fig ) . Household level unhygienic practices were more common among households with infrastructure that was suspected to increase cholera risk . For example , households having no hand soap at home tended to tended to share a latrine with neighbors ( 1 . 11 , 95%CI = 1 . 02 , 1 . 19 ) ( Fig 4A ) . Moreover , multiple risk factors related to water supply and sanitation infrastructure tended to exist in the same households; sharing a latrine was more common among households that were over 10m to the nearest drinking water source ( 1 . 04 , 95%CI = 1 . 00 , 1 . 08 ) ( Fig 4A ) . In addition , among the same matched-sets , unhygienic practices were correlated with infrastructure that was suspected to increase cholera risk . For example , households not always boiling drinking water tended to be in the same matched-set as those using a pit latrine ( 1 . 21 , 95%CI = 1 . 07 , 1 . 36 ) and those over 10m to the nearest water source ( 1 . 16 , 95%CI = 1 . 07 , 1 . 35 ) ( Fig 4B; See S2 Table for co-occurrence of exposures with not boiling tubewell water and not boiling supplied water ) . Furthermore , the use of tubewell water and distance to water source were correlated in the same matched-sets ( 1 . 22 , 95%CI = 1 . 06 , 1 . 48 ) ( Fig 4B ) . Clustering between individual level exposures appeared to mostly reflect the differing domestic roles of study participants , although the results were mostly statistically insignificant ( S2 Fig ) .
We found substantial clustering of single exposures and co-occurrence of different exposures suspected to modify cholera and diarrhea risk within groups of nearby households , and an overall trend of households nearer to one another being more likely to share risk factors . This suggests that even within a relatively compact neighborhood there can be substantial spatial dependence of risk for cholera and diarrheal diseases driven by spatial homogeneity in behavior or access to improved water and sanitation . Aggregating risk profiles at the neighborhood-level may be misleading and obfuscate hidden spatial heterogeneity , hence we may miss pockets of high risk . Thus , interventions targeted at micro-scale clusters of high cholera risk may be an important complement to general disease prevention or control campaigns . In cholera endemic neighborhoods , clusters of high-risk behavior could be identified prospectively through community surveys , or by canvasing the area around incident cholera cases . These groups could then be slated for more intensive water , sanitation , and hygienic interventions including education and provision of supplies . This latter course has proven effective in preventing onward spread of cholera in households , and may work equally well within clusters with high cholera risk [24] . Some study designs must account for spatial clustering of exposures at a fine scale . In a case control study , matching controls by proximity aims to control for environmental confounders that are unknown or difficult to measure , and is an efficient and practical method for risk factor studies or vaccine effectiveness studies in low-resource settings [25 , 26] . However , if we do not increase sample size to account for the correlation in risk , spatially clustered designs will be underpowered to detect the effect of even non-clustered exposures , and this form of overmatching cannot be corrected in the analysis phase [8] . In our study , use of supplied water showed significant clustering within matched-sets , and hence was effectively matched on . If contaminated supplied water was associated with risk of cholera in Arichpur , we would be underpowered to detect a significant association unless we increased the sample size . Another study design , the cluster sample survey , has been widely used for estimating vaccination coverage and disease burden . In each cluster , households are usually selected based on their proximity to index households to reach the designated sampled size [27] . Clustering of risk factors increases design effect and our study has shown that it could be profound within a small community . The study has several limitations . Our analysis focused only on one neighborhood in urban Bangladesh . Factors that are specific to the Arichpur neighborhood such as the network of water infrastructure affect the spatial scale of clustering . However , the result may likely be observed in other densely populated urban areas where cholera is most prevalent and similar risk profile is present . Thus , a clear understanding of risk factors present in the community is important for cholera interventions and studies . In addition , some exposures in our study are self-reported . Participants may be more likely to report low risk behavior , and may have undermined the level of clustering . However , the conclusion of microscale clustering would hold especially for the household level exposures , because they were less likely to be biased as infrastructure related to water supply and sanitation are very unlikely to change . Furthermore , 11 out of the 69 hospital-based cholera positive households did not participate in our study , and 13 out of the 31 households recruited from the pharmacy did not provide stool samples ( Fig 1 ) . If these missing households systematically differ from those in the study , our results may be biased . In conclusion , our study demonstrates clustering of individual and groups of exposures modifying risk of cholera and diarrheal diseases within immediate neighbors , and an overall trend of clustering over the entire Arichpur neighborhood . Based on our findings on clustering of high-risk behaviors , we recommend incorporating community risk assessments to supplement general water , sanitation , and hygienic intervention campaigns . For clustered studies such as those designed to assess vaccine effectiveness and identify risk factors , we call for an initial risk assessment for level of risk factor clustering , and determine the required sample size accordingly . | While clustering of cholera incidence had been previously described , the relative role of similar risk behaviors versus transmission dynamics is not well understood . We explored how risk factors for cholera clustered at the sub-community scale , and found significant more correlation in risk behaviors among spatially matched households than the community as a whole . We found clustering of single risk factors , and co-occurrence of different risk factors . Our results indicated that the distribution of risk behaviors may play a role in the clustering of cholera cases at very small ( e . g . , <100m ) spatial scales . This had implications for spatially matched study designs , which may be overmatching on some exposures . It also may lead us to rethink targeted interventions , suggesting a role for more intensive highly targeted interventions as a supplement to more generalized campaigns . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"water",
"resources",
"tropical",
"diseases",
"condensed",
"matter",
"physics",
"geographical",
"locations",
"health",
"care",
"bacterial",
"diseases",
"physiological",
"processes",
... | 2016 | Micro-scale Spatial Clustering of Cholera Risk Factors in Urban Bangladesh |
RNA–directed DNA methylation ( RdDM ) is an epigenetic control mechanism driven by small interfering RNAs ( siRNAs ) that influence gene function . In plants , little is known of the involvement of the RdDM pathway in regulating traits related to immune responses . In a genetic screen designed to reveal factors regulating immunity in Arabidopsis thaliana , we identified NRPD2 as the OVEREXPRESSOR OF CATIONIC PEROXIDASE 1 ( OCP1 ) . NRPD2 encodes the second largest subunit of the plant-specific RNA Polymerases IV and V ( Pol IV and Pol V ) , which are crucial for the RdDM pathway . The ocp1 and nrpd2 mutants showed increases in disease susceptibility when confronted with the necrotrophic fungal pathogens Botrytis cinerea and Plectosphaerella cucumerina . Studies were extended to other mutants affected in different steps of the RdDM pathway , such as nrpd1 , nrpe1 , ago4 , drd1 , rdr2 , and drm1drm2 mutants . Our results indicate that all the mutants studied , with the exception of nrpd1 , phenocopy the nrpd2 mutants; and they suggest that , while Pol V complex is required for plant immunity , Pol IV appears dispensable . Moreover , Pol V defective mutants , but not Pol IV mutants , show enhanced disease resistance towards the bacterial pathogen Pseudomonas syringae DC3000 . Interestingly , salicylic acid ( SA ) –mediated defenses effective against PsDC3000 are enhanced in Pol V defective mutants , whereas jasmonic acid ( JA ) –mediated defenses that protect against fungi are reduced . Chromatin immunoprecipitation analysis revealed that , through differential histone modifications , SA–related defense genes are poised for enhanced activation in Pol V defective mutants and provide clues for understanding the regulation of gene priming during defense . Our results highlight the importance of epigenetic control as an additional layer of complexity in the regulation of plant immunity and point towards multiple components of the RdDM pathway being involved in plant immunity based on genetic evidence , but whether this is a direct or indirect effect on disease-related genes is unclear .
RNA-directed DNA methylation ( RdDM ) is an epigenetic modification mechanism driven by noncoding small interfering RNAs ( siRNAs ) [1] , [2] . siRNAs are present in most eukaryotic organisms , are highly developed in plants and regulate gene expression at the transcriptional and posttranscriptional level in a sequence-specific manner . In contrast to microRNAs ( miRNAs ) that are derived from the transcripts of miRNA genes generated by RNA Polymerase II , production of RdDM-associated siRNAs requires RNA Polymerase IV ( Pol IV ) complex activity which includes , among other constituents , the largest and second largest subunits , NRPD1 and NRPD2 , respectively [3]–[5] . Upon the action of Pol IV , the resulting single-stranded RNAs are used as templates for RNA-dependent RNA polymerase 2 ( RDR2 ) generating double-stranded RNAs , which are processed by DICER-LIKE 3 ( DCL3 ) [6] , [7] . Subsequently , RNA methyltransferase HUA ENHANCER-1 ( HEN1 ) generates functional siRNAs that are recruited by ARGONAUTE4 ( AGO4 ) to form the AGO4-RISC multiprotein complex guided to siRNA-complementary genome sequences [8]–[10] . AGO4-siRNA complexes interact with the RNA Polymerase V ( Pol V ) complex , which includes the largest and second largest subunits , NRPE1 and NRPD2 , respectively . Pol V is somehow required to recruit DRM2 methyltransferase as well as histone-modifying complexes to finally establish the methylation pattern in the siRNA-complementary genome sequences; however , the details of this recruitment are unknown . This process results in the methylation of certain genome repeat regions and their subsequent transcriptional silencing [2] . Among the different classes of siRNA , the 24 nt in lenght hetrocromatic siRNAs ( hc-siRNAs ) and repeat-associated siRNAs ( ra-siRNAs ) , primarily derived from transposons , repeated elements and heterochromatin regions , are those functioning in the RdDM pathway by mediating DNA methylation and/or histone modification at the target sites [2] . Small RNAs regulate a multitude of biological processes in plants , including sustaining genome integrity , development , metabolism and responses to changing environmental conditions and abiotic stress [11] . Increasing evidences also indicate that plant endogenous small RNAs , including miRNAs and siRNAs , are integral regulatory components of plant defense machinery against microbial pathogens [12] . The Arabidopsis miR393 imparts basal resistance to the bacterial pathogen Pseudomonas syringae DC3000 by targeting the auxin receptors TIR1 , ABF2 and ABF3 [13] . Besides miR393 , two other miRNA families , miR160 and miR167 , are upregulated following PsDC3000 inoculation and target members of auxin-response factors ( ARF ) [14] . Thus , in response to bacterial infection , plants suppress multiple components of the auxin signaling pathway . In turn , bacteria have developed type III secretion effectors that repress transcription of miRNA genes , the host RNA silencing machinery is suppressed and therefore disease susceptibility increase [15] . Similarly , Lu et al . [16] identified a series of 10 miRNAs families in loblolly pine whose expression were suppressed , and the transcript levels of their target genes increased , upon infection with the rust fungus Cromartium quercuum f . sp . fusiforme . Likewise , upon infection of Brasica rapa with Turnip mosaic virus ( TuMV ) the miR1885 is upregulated , and its target is predicted to be a member of the TIR-NBS-LRR class of disease-resistance proteins [17] . Thus , it appears that following detection of pathogen-associated molecules , plant cells undergo changes in miRNA global profiles that mediate the establishment of a specific defense response [12] , [18] . Although plants contain only several hundred miRNAs , they contain huge numbers of endogenous siRNAs but only in a few cases the involvement of siRNAs in plant immunity has been described . In Arabidopsis , the natural antisense transcript ( NAT ) -derived nat-siRNAAATGB2 and the long siRNA lsiRNA-1 , which specifically targets the mitochondrial pentatricopeptide protein ( PPR ) -like gene PPRL and the RNA-binding protein gene AtRAP , respectively , are induced by the bacterial pathogen PsDC3000 ( avrRpt2 ) and contribute to plant antibacterial immunity [19] , [20] . The endogenous siRNAs generated at disease resistance RPP4 locus , which impart resistance to both the bacterial Ps pv . maculicola and the oomycetes Hyaloperonospora arabidopsidis , constitute a third example for siRNA-mediated resistance responses [21] . However , it remains unclear how RdDM participates in this type of processes . The understanding of the overall contribution and requirement of the different components that conform the RdDM pathway , and how important they are in the regulation of the RdDM-mediated processes , particularly in relation to plant immunity , is an issue that still remains to be fully understood . Previously we described a genetic screen in Arabidopsis design to identify mutants ( ocp mutants ) with altered immune responses [22] . This allowed identifying AGO4 , through the characterization of its mutant allele ago4-2/ocp11 , as an important component of the RdDM pathway in mediating plant immune responses towards PsDC3000 [23] . Towards characterizing the contribution of other components of the RdDM pathway in plant immunity , we report here on the isolation and characterization of ocp1 , a recessive mutant allele of NRPD2 . Our results support that RdDM , through the action of RNA Pol V , is pivotal in modulating immune responses towards pathogens .
The Arabidopsis ocp mutants were identified previously in a genetic screen [22] designed to isolate negative regulators of pathogen-induced defense responses . The H2O2-responsive and defense-related Ep5C gene promoter fused to GUS was used as reporter [24] . Here we described the characterization of the ocp1 mutant . Figure 1A shows the constitutive Ep5C::GUS expression in rosette leaves from ocp1 plants compared with its parental Col-0 line ( line 5 . 2 ) . ocp1 plants exhibited similar plant architecture and growth habit to the wild-type plants ( Figure 1B ) . F1 hybrids from a backcross between parental and ocp1 plants showed the absence of GUS activity , and GUS activity segregated in the F2 progeny as a single recessive Mendelian locus [OCP1:ocp1 , 111:33 ( P<0 . 05 , χ2 test ) ] . We hypothesize that the constitutive expression of Ep5C::GUS observed in ocp1 plants might be accompanied by an altered disease resistance response to pathogens as previously revealed in ocp3 and ocp11 plants [22] , [23] , [25] , [26] . Therefore , we inoculated ocp1 plants with the virulent necrotrophic fungal pathogen Botrytis cinerea and monitored the disease response in leaves in comparison with the parental line . Disease was scored by recording the extent of necrosis . Wild-type plants exhibited normal susceptibility to B . cinerea ( Figure 1C ) , with inoculated leaves showing necrosis accompanied by extensive proliferation of the fungal mycelia . In contrast , ocp1 plants showed increased susceptibility to B . cinerea distinguished by moderate but statistical significant enlargement of necrotic areas at inoculation sites ( Figure 1C ) . Susceptibility of ocp1 plants to pathogens was also investigated with the bacterial pathogen PsDC3000 . The npr1 mutant , which is compromised in resistance towards this pathogen [27] was used as a control . Resulting bacterial growth in inoculated leaves is shown in Figure 1D and indicates the wild-type and ocp1 mutant susceptibility was unchanged towards virulent PsDC3000 . In addition , plants were inoculated with an avirulent strain of PsDC3000 carrying the avrRpm1 gene that triggers a hypersensitive cell death response in the plant that stops bacterial growth . The rpm1 mutant , compromised in the hypersensitive response and consequently hypersusceptible to the pathogen , was used as a control . Results showed the growth of PsDC3000 ( avrRpm1 ) in ocp1 plants was not significantly different to that observed in wild-type plants ( Figure 1E ) . These results were consistent with normal accumulation of transcripts of the salicylic acid ( SA ) -responsive gene PR-1 at 48 h following inoculation with PsDC3000 ( Figure 1F ) , however induction occurs earlier in ocp1 plants . Interestingly , induction of the jasmonic acid ( JA ) -responsive gene PDF1 . 2a , a characteristic molecular response of plants to fungal attack , was compromised in ocp1 plants following inoculation with B . cinerea ( Figure 1G ) . This later observation is congruent with the observation that ocp1 plants showed enhanced disease susceptibility to this pathogen ( Figure 1C ) . The genetic lesion carried by ocp1 plants was identified by positional cloning ( Figure S1 ) . A single nucleotide deletion was detected on locus At3g23780 , particularly in the third exon of the transcribed gene encoding NRPD2 , the second largest subunit of the RNA Pol IV and Pol V protein complexes ( Figure 2A and Figure S1C ) . The loss of a nucleotide residue created a change in the NRPD2 open reading frame that leads to a frame shift starting at residue 595 ( Figure 2A ) followed by an incorrect 22 amino acid C-terminal tail sequence before an in-frame stop codon ( Figure S2 ) . The mutation renders a protein of 616 amino acid residues , instead of the 1172 contained in NRPD2 , that thus has lost almost half of the protein sequence , including the amino acids that contribute to the active site of RNA polymerases [28] . The result obtained in our mapping strategy was corroborated with a test of allelism between ocp1 plants and plants carrying a null allele of NRPD2 , in particular with nrpd2-2 plants which carry a T-DNA insertion ( SALK_046208 ) [3] . Analysis of GUS expression driven by the Ep5C gene promoter in 20 F1 plants derived from a cross between homozygous ocp1 plants with homozygous nrpd2-2 plants or , alternatively , from a reversed cross between nrpd2-2 plants with ocp1 plants , revealed that all F1 plants showed constitutive GUS expression ( Figure S3 ) . Conversely , control crosses between the parental Col-0 plants carrying the Ep5C::GUS gene construct ( line 5 . 2 ) with either ocp1 plants or nprd2-2 plants revealed no GUS expression in any of the F1 22 plants analyzed ( Figure S3 ) . The result indicates that the ocp1 and nrpd2-2 are mutant alleles of the same NRPD2 gene and supported the conclusion that the ocp1 mutation represents a loss of function allele . Hence , the ocp1 mutation will be referred also as ocp1/nrpd2-53 . From the type of mutation found , we cannot exclude the possibility that ocp1 plants are still able to produce a truncated version of the NRPD2 protein with a residual ability to interact with other components of the RNA polymerase complexes . Since Pol IV and Pol V complexes are comprised of a variety of interacting subunits , some being polymerase-specific while other subunits shared ( i . e . , NRPD2 ) [5] , [29] , [30] , and with some cross-talk described for some of their subunits ( i . e . , between NRPD2 and NRPE1; [4] ) , we can not discard the possibility that the relationships between the different components of the two RNA polymerase complexes may become differentially altered in the ocp1 mutant . In this respect , the availability of the ocp1 allele may represent a valuable experimental tool to approach the biochemical regulation of the RdDM mechanism . Interestingly , RT-PCR analyses of NRPD2 transcript levels in ocp1 plants revealed the absence of notable changes in gene expression compared with Col-0 plants ( Figure 2B ) . This is in marked contrast with the expression observed in nrpd2-2 null mutant plants where no transcript amplification products can be obtained ( Figure 2B ) . A comparison of the disease resistance response between ocp1 and nrpd2-2 plants revealed that while the ocp1 plants showed a moderate increase in susceptibility to B . cinerea , the nrpd2-2 null mutant responded to B . cinerea infection with a remarkable enhancement in susceptibility ( Figure 2C ) . The enhanced susceptibility phenotype of nrpd2-2 plants was further corroborated by recording the susceptibility towards Plectosphaerella cucumerina , a different fungal necrotroph ( Figure 2D ) . Consistent with the observed increase in disease susceptibility to P . cucumerina , RT-qPCR experiments revealed that induction of the JA-responsive PDF1 . 2a gene was disabled in nrpd2-2 plants compared to Col-0 ( Figure 2E ) . These results mirror what occurs in ocp1 plants following B . cinerea infection ( Figure 1F ) . Of importance for understanding the immune-related phenotype of nrpd2-2 plants is the observation that expression of the SA-responsive PR-1 gene was clearly enhanced following fungal inoculation in the mutant when compared to wild-type plants ( Figure 2F ) . Since nrpd2-2 plants show an enhanced disease susceptibility of bigger magnitude than that observed in ocp1/nrpd2-53 plants , subsequently , the experiments related to disease resistance/susceptibility will be carried out employing the nrpd2-2 allele . To further substantiate the molecular phenotype of ocp1 plants in relation to RdDM , we checked if the methylation status of different RdDM target sequences could be similarly affected in ocp1 and nrpd2-2 plants . We analyzed the methylation status in ocp1 plants of the RdDM pathway DNA target sequences SUPERMAN , ribosomal 5S genes and the retrotransposon AtSN1 [31] . We used methylation tests employing the methylation-sensitive restriction endonuclease HaeIII ( where HaeIII will not cut DNA if methylated ) , with subsequent amplification by PCR [32] . Initial experiments revealed that ocp1 , as well as ago4-2/ocp11 plants used as controls , exhibit a higher degree of hypomethylation in SUPERMAN gene compared to Col-0 plants ( Figure 3A ) . Analyses were extended to the ribosomal 5S genes and the AtSN1 retrotransposon and we incorporated nrpd2-2 , nrpd1-3 and nrpe1-1 mutants for comparison . Figure 3B shows mutants demonstrated higher degrees of hypomethylation in the sequences analyzed . DNA samples derived from ocp1 plants exhibited decreased amplification for the 5S and AtNS1 loci , confirming a clear DNA methylation deficiency in this mutant . The ABI5 gene , whose sequence contains no restriction sites for HaeIII , was used as a control . Methylation tests were also used to ascertain whether or not the enhanced induction observed for the PR-1 gene , or the repression of PDF1 . 2a , in the nrpd2 mutant following fungal infection correlated with defects in the DNA methylation of their promoter regions . Since both genes contain a large number of recognition sites for the methylation-sensitive restriction enzymes FspEI , MspJI and AvaII ( 168 target sequences in the PR-1 gene and 298 targets in the PDF1 . 2a gene ) , and where FspEI and MspJI sites must be methylated for the enzymes to cleave the DNA , we used restriction analysis with these enzymes with subsequent amplification by PCR to check the methylation status of the PR-1 and PDF1 . 2a genes . The results shown in Figure S4 and Figure S5 revealed that none of the promoters appear methylated , not even in Col-0 plants . Conversely , the sensitivity of the methylated 5S ribosomal DNA ( Figure S4 ) to the aforementioned enzymes revealed the appropriateness of the method used to identify methylation of cytosine residues . The lack of a methylation footprint in the DNA of the defense-related PR-1 and PDF1 . 2a genes might suggest that the abnormal expression patterns concurring in nrpd2 mutant plants must obey not to a direct modification of cytosine residues but to other type of chromatin modification or mechanism similarly controlled either directly or indirectly by the RdDM pathway . As for NRPD2 , we addressed if other RdDM pathway components are similarly engaged in plant immunity . A comparative analysis of the disease resistance response of nrpd1 , nrpe1 , and ago4 mutant plants due to inoculation by B . cinerea was performed in relationship to nrpd2 . Figure 4A shows an increase in nrpe1 disease susceptibility to B . cinerea; the susceptibility being of a magnitude similar to that attained in nrpd2 plants . This enhancement in susceptibility was comparatively greater than that observed in ocp1 plants but less than in ago4-2/ocp11 plants . Conversely , nrpd1 plants did not exhibit a significant deviation from the normal disease response observed in Col-0 plants . This differential behavior was further corroborated in the Pol IV and Pol V defective mutants by challenging with P . cucumerina ( Figure 4B ) . The nrpd1 nrpe1 double mutant that would be defective in both Pol IV and Pol V activities was incorporated in this experiment for comparison . nrpd1 nrpe1 plants showed an enhanced disease susceptibility of a magnitude similar to that attained in nrpe1 or nrpd2 plants . Furthermore , fungal biomass determination in leaves inoculated with P . cucumerina , as an alternative method for recording disease resistance , also revealed that the single nrpd2 and nrpe1 mutants , as well as the double nrpd1 nrpe1 mutant support significantly more fungal growth than Col-0 and the nrpd1 mutant ( Figure S6 ) . Therefore , the Pol V complex participates in the regulation of the immune response to necrotrophs while the Pol IV complex appears at least partially dispensable . This is sustained also by the observation that expression patterns of the PDF1 . 2a and the PR-1 genes in nrpd1 plants are dissimilar to that commonly attained in both nrpe1 and nrpd2 plants , either in the course of infection with P . cucumerina ( Figure 4C–4D ) or upon chemical induction by treating plants with a solution of either 0 . 5 mM SA ( Figure S7A ) or 0 . 1 mM JA ( Figure S7B ) . Notorious is the higher JA-triggered PDF1 . 2 gene induction in nrpd1 plants in comparison to Col-0 ( Figure S7B ) . Conversely , this JA-triggered PDF1 . 2a gene induction is notably repressed in the Pol V defective mutants ( Figure 4C and Figure S7B ) . This is in marked contrast with the altered expression pattern observed in nrpd2 plants where induction of PR-1 gene expression showed enhancement following inoculation with P . cucumerina ( Figure 4D ) or upon external application of SA ( Figure S7A ) . Importantly , this pattern of gene expression was reproduced in nrpe1 plants ( Figure 4D and Figure S7A ) . Moreover , the transcription factors WRKY6 and WRKY53 that bind W-box and transcriptionally regulate gene expression of SA-related genes , including PR-1 [33] , and are themselves induced by pathogen infection [34] , show similar enhanced level of induction following SA application in nrpd2 and nrpe1 plants when compared to Col-0 or nrpd1 plants ( Figure S7C and S7D ) . To further extent the implication of RdDM mechanism in plant immunity we inoculated the drm1drm2 double mutant plants , which is compromised in de novo DNA methylation [35] , with P . cucumerina and recorded the disease response . Results in Figure 4E reveal that loss of the functional RDM methyltransferase compromise disease resistance and results in plants showing enhanced susceptibility to P . cucumerina to levels even higher than in nrpd2 plants . This reinforces the proposal that RdDM is pivotal for plant immunity . Likewise , we observed that the chromatin-remodeling factor DRD1 , which is required for the association of NRPE1 with chromatin [36] , is also pivotal for plant immunity . Figure 4E shows that drd1-6 plants phenocopy Pol V defective mutants . The RNA-dependent RNA polymerase 2 ( RDR2 ) , which functions early in the RdDM pathway by generating dsRNA from the ssRNA transcripts thought to emanate presumably from the the Pol IV complex was also entertained in these experiments . Intriguingly , rdr2 plants also show a strikingly enhancement in susceptibility to fungal infection ( Figure 4E ) , achieving highest levels of susceptibility to P . cucumerina . This observation strongly argues in favor of RDR2 as required for plant immunity . Then , how can a mechanism explain the exclusion of RNA Pol IV in mediating plant immunity while the rest of downstream components of the RdDM pathway are engaged in this biological process ? There are previous evidences where Pol V has been described to operate independently of Pol IV , such as in the mechanism for maintaining the methylation status of target sequences [37] , and thus for some processes Pol IV and Pol V act not in concert [38] . A hypothesis that could help explain the paradoxical observations indicating that Pol V , RDR2 , AGO4 , DRD1 and DRM2 , but not Pol IV , are required for plant immunity to fungal pathogens could be one where RDR2 can accept RNA transcripts derived from the action of RNA Pol V , and not necessarily only from RNA Pol IV . These putative transcripts thought to be acted upon by RDR2 , which generates dsRNA , will be processed into siRNAs and feed into the RdDM pathway . In support for the existence of Pol V-dependent transcripts required for DNA methylation and silencing is the recent identification of low-abundance intergenic non-coding ( IGN ) transcripts [36] . It could be that a similar situation is on the basis to explain the requirement of RdDM for plant immune responses . This possibility merits future research approaches . The previous results suggest that in Pol V defective mutants SA-related defense genes are poised for enhanced activation following perception of pathogenic cues and concurrently JA-related defenses appear impeded for induction . This will be congruent with a notion where Pol V may regulate a priming phenomenon for SA-mediated defense responses that ultimately would modulate the speed and extent of gene activation . However , the lack of a methylation footprint in the DNA of the defense-related PR-1 and PDF1 . 2a genes ( Figure S5 and Figure S6 ) suggest that the observed abnormal gene expression patterns concurring in the Pol V defective mutants is not to be due to an altered DNA methylation pattern resulting from a defective RdDM pathway . However , one could still entertained the possibility that changes in chromatin structure such as those obeying to covalent modification of histones , which are also under the control of the RdDM pathway , may be on the basis for the enhanced expression observed for PR-1 and , therefore , for the altered resistance phenotypes in the mutant plants . This would be congruent with the recent identification of a mechanism linking chromatin modification in wild type plants , through the differential modification of histones in several genes encoding WRKY transcription factors ( i . e . WRKY6 , WRKY29 or WRKY53 ) , with priming of a defense response following pharmacological treatment with the SA analogue acidobenzolar S-methyl ( BTH ) which functions as a priming agent in plants [39] . Thus , we hypothesized that in Pol V defective mutants PR-1 could be poised for enhanced activation of gene expression by a differential modification of histones . By using chromatin immunoprecipitation ( ChIP ) we analyzed trimethylation of histone H3 Lys4 ( H3K4me3 ) and acetylation of histone H3 Lys9 ( H3K9ac ) on the promoter of the PR-1 gene . For comparison , the promoter of the JA-inducible PDF1 . 2a gene , that of the constitutively expressed Actin2 gene and also those of the WRKY6 and WRKY53 genes were similarly studied . The specificity of the ChIP reaction was evaluated in advance by measuring histone modifications on these genes in Col-0 plants treated with BTH ( Figure S8A and S8B ) . On the PR-1 promoter H3K4me3 and H3K9ac marks increased after BTH application while these marks did not change in the promoters of Actin2 or PDF1 . 2a ( Figure S8A and S8B ) . As for PR-1 , these chromatin marks were similarly increased in the promoters of WRKY6 and WRKY53 upon treatment with BTH ( Figure S8C ) . Thus chromatin marks normally associated with active genes [39] , [40] are set in the promoters of SA-related defense genes by the priming stimulus of BTH . Interestingly , determination of H3K4me3 ( Figure 5A ) and H3K9ac ( Figure 5B ) chromatin marks in the PR-1 promoter in ChIP samples derived from nrpd2 and nrpe1 plants , revealed that these marks are already set in these two mutants , although PR-1 gene activation does not take place . Thus , Pol V defective mutants mimic Col-0 plants treated with the priming agent BTH . This reconciles with the idea that the PR-1 gene is switch on for priming in the Pol V defective mutant and explains why this gene shows enhanced induction upon pathogenic attack in the same mutants ( Figure 4D ) . In the nrpd1 mutant only a moderate increase in the setting of these chromatin marks in the promoter of PR-1 was detected ( Figure 5A and 5B ) . No variation in similar activation marks was observed in the promoters of the Actin2 and PDF1 . 2a genes ( Figure 5A and 5B ) . Other histone marks , such as H3K9me2 and H3K27me3 , both of which repressive marks normally associated with heterochromatin and established through the RdDM pathway [41] , appear notably reduced in the PR-1 promoter in ChIP samples derived from nrpd2 and nrpe1 plants , and much less reduced in nrpd1 plants , when compared to Col-0 plants ( Figure S9A and S9B ) . Moreover , Col-0 plants respond to P . cucumerina infection with reduction in the setting of these two repressive histone marks in the PR-1 gene promoter but not in the promoters of the PDF1 . 2a or Actin2 genes ( Figure S9C ) . The dismantling of histone repressive marks in infected plants , along with the concurring increase in histone activation marks and decrease in repressive marks in the promoter of the PR-1 gene , as observed in nrpd2 and nrpe1 plants , gives further support to the implication of Pol V in regulating defense gene activation . As for PR-1 , H3K4me3 activation marks are also constitutively set in the promoters of the WRKY6 and WRKY53 genes in healthy nrpd2 and nrpe1 plants ( Figure S8C ) , again mirroring the effect carried out by BTH on Col-0 for these promoters ( Figure S8C ) . Further analysis demonstrated that Col-0 plants respond to P . cucumerina infection with a drastic increase in the setting of H3K4me3 and H3K9ac activation marks in the promoters of PR-1 ( Figure 5C and 5D ) . In nprd2 plants , in which these chromatin marks are already set in PR-1 , P . cucumerina inoculation further increase H3K4me3 marks on the PR-1 promoter to levels that are even higher than those attained in Col-0 ( Figure 5C ) . However , for H3K9ac marks no further increase was observed in nrpd2 plants , suggesting that this type of mark is completely set in the mutant . In contrast , no variation in the setting of these chromatin marks was detected in the PDF1 . 2a promoter upon fungal infection ( Figure 5C and 5D ) . For WRKY6 and WRKY53 gene promoters , Col-0 plants respond to P . cucumerina infection by similarly increasing H3K4me3 mark setting in both promoters ( Figure S10 ) . Compared to Col-0 , nrpd2 plants constitutively carry increased H3K4me3 mark setting in WRKY6 and WRKY53 gene promoters and do not show further increases upon inoculation , but instead slightly decrease ( Figure S9 ) . Together , these data imply that Pol V , either directly or indirectly , regulates the extent of chromatin modifications on SA defense-related gene promoters , and may be the underlying mechanism controlling priming marks facilitating the more rapid activation of gene expression observed upon perception of pathogenic cues . As reported for other genes , the observed covalent modifications in chromatin might provoke increases in the accessibility of DNA or perhaps in the provision of docking sites for gene activators [42] , [43] . Enhanced activation of SA-mediated defenses is characteristic of plants resistant to biotrophic pathogens , like PsDC3000 , and is on the basis for a systemic type of immunity known as systemic acquired resistance ( SAR ) [44] . Our results on a priming effect for enhanced expression of SA defense-related genes in nrpd2 and nrpe1 plants suggest these mutants may be altered in the resistance to PsDC3000 . Consequently , we addressed Pol IV and Pol V defective mutants in search for defects in the immune response to PsDC3000 . We used ago4-2/ocp11 and npr1 plants as controls , both exhibiting heightened PsDC3000 disease susceptibility [23] , [27] . Interestingly , a significant enhanced disease resistance to PsDC3000 was observed in nrpd2 , nrpe1 , and in nrpd1 nrpe1 plants , when compared to Col-0 plants ( Figure 6 ) . In contrast , statistically significant effects were not observed in nrpd1 plants relative to Col-0 in response to PsDC3000 , giving further support to the idea that RNA Pol IV seems not engaged in plant immunity . The observed heightened resistance towards PsDC3000 in nrpd2 and nrpe1 plants indicated that in wild-type plants Pol V is required for susceptibility to this pathogen . However , in ago4-2/ocp11 plants resistance to PsDC3000 is severely compromised . Although there is no obvious explanation for this contrasting effect , as previously stated [23] one can speculate that AGO4 can serve a novel function , and while required for an effective defense response it may operate independently of the RdDM pathway . An important observation derived from the results presented is the co-existence of an enhanced disease resistance to a biotrophic bacteria , like PsDC3000 , with an enhanced susceptibility to necrotrophic fungi in Pol V defective mutants . This reveals an underlying complexity in the control of disease resistance by RdDM . The SA and JA signal pathways are under an antagonistic equilibrium that occasionally culminates with the partial inhibition of one pathway when the other is facilitated . Consequently the interaction between pathways serves to optimize responses to a specific type of pathogenic insult [45] . Our results demonstrated that nrpd2 and nrpe1 plants are poised for enhanced activation of SA defense-related genes and respond to pathogen attack with a marked enhancement in the induced expression of marker genes , which suggests these plants are more prone to mobilize the defense arsenal controlled by SA . A simpler explanation for these observations is that in wild type plants Pol V negatively regulates a priming mechanism for SA-mediated disease resistance while keeping intact a JA-mediated disease resistance . Defects in Pol V function , such as those observed in nrpd2 and nrpe1 mutants , de-repress the priming mechanism for SA-mediated resistance through pertinent chromatin modifications , and renders enhanced resistance to PsDC3000 . As a tradeoff , presumably mediated through endogenous antagonistic cross talk mechanisms , mis-regulation of the JA-mediated disease resistance occurs . This thus explaining the repressed expression of JA-marker gene and the heightened susceptibility of nrpd2 and nrpe1 plants to fungal pathogens . However , although this mechanism seems very likely , we still cannot disregard the possibility that RdDM may be similarly required for normal expression of one or more unknown genes involved in JA signaling . Disruption of RdDM could thus lead to a disruption of JA signaling which would in turn result in hyper-activation of SA signaling . In fact , mutant plants with JA-mediated signaling pathway defects and hypersensitivity to fungal necrotrophs concurrently present a less repressed SA-mediated signaling pathway , resulting in a more efficient defense response when challenged with biotrophic pathogens [45] , [46] . Experiments directed towards identification of an epigenetic footprint associated to the JA pathway merits future reach and will help clarify the complexity of the antagonistic cross-talk mechanism between the SA and the JA signal transduction pathways . A deeper understanding on how the RdDM and associated chromatin modification acts as a mechanism controlling gene priming and induced immune responses in plants , and how pathogens may counteract this epigenetic regulation for their own benefit will open new avenues for the a better knowledge on how plant immunity is orchestrated .
Arabidopsis were grown in a growth chamber ( 19 to 23°C , 85% relative humidity , 100 µE m−2 s−1 fluorescent illumination ) under a 10/14 h light/dark photoperiod . All mutants are in Col-0 background . ago4-2/ocp11 , npr1 , rpm1-1 , rdr2 , drd1-6 and drm1/drm2 plants were previously described [23] . nrpd2-2 ( SALK_046208 ) ; nrpe1-11 ( SALK_029919 ) and nrpd1-3 ( SALK_128428 ) were obtained from the Salk Institute Genomic Analysis Laboratory ( http://signal . salk . edu/ ) . nrpd1 nrpe1 double mutant was obtained from T . Lagrange . Plant leaves were incubated overnight at 37°C in GUS staining buffer as previously described [22] . The ocp1 mutant was backcrossed twice to the PEp5C:GUS line to confirm its recessive inheritance . ocp1 plants were crossed to Ler , and F1 plants were allowed to self . F2 plants were scored for co-segregation of high constitutive GUS activity with simple sequence length polymorphisms ( SSLP ) [40] . Molecular markers were derived from the polymorphism database between the Ler and Col-0 ecotypes ( http://www . arabidopsis . org ) . Methylation tests using the methylation-sensitive endonuclease HaeIII , FspEI , AvaII and MspJI were performed as described [32] . The relative DNA fragment amounts corresponding to SUPERMAN , 5S and AtSN1 were obtained after 30 , 25 and 35 respective PCR cycles . For ABI5 , 30 ( A ) or 26 ( B ) PCR cycles were used . PR-1 and PDF1 . 2a methylation assays are provided in a supplemental file . Gene expression analysis , by either RT-PCR or RT-qPCR was performed as described previously [23] . The primers used to amplify the different genes and DNA regions , and the PCR conditions employed for genotyping T-DNA insertions , and RT-PC and qRT-PC experiments are provided in the supporting information file Text S1 . Bacterial strains were grown overnight and used to infect 5-week-old Arabidopsis leaves by infiltration and bacterial growth determined as described [23] . Twelve samples were used for each data point and represented as the mean ± SEM of log c . f . u . /cm2 . B . cinerea and P . cucumerina bioassays were performed as previously described [24] . Fungal disease symptoms were evaluated by determining the lesion diameter ( in mm ) of a minimum of 30 lesions . All experiments were repeated at least three times with similar results . Chromatin isolation and immunoprecipitation were performed as described [47] . Chip samples , derived from three biological replicates , were amplified in triplicate and measured by quantitative PCR using primers for PR-1 , WRKY6 , WRKY53 and Actin2 as reported [39] . The rest of primers are described in Text S1 . All ChIP experiments were performed in three independent biological replicates . The antibodies used for immunoprecipitation of modified histones from 2 g of leaf material were antiH3K4m3 ( #07-473 Millipore ) , antiH3K4ac ( #07-352 Millipore ) , antiH3K9me2 ( ab1772 Abcam ) and antiH3K27me3 ( ab6002 Abcam ) . | The influence of epigenetic regulation in controlling the adaptive responses of living organisms to changes in the environment is becoming a common theme in biology . RNA–directed DNA methylation ( RdDM ) is an epigenetic control mechanism driven by a subset of noncoding small interfering RNAs ( siRNAs ) that influence gene function without changing DNA sequence by inducing de novo methylation of cytosines , or by modification of histones , at their target genomic regions . The implication and roles of the RdDM mechanism in the orchestration of plant immune responses still remains to be characterized . A recent study in the model plant Arabidopsis showed that ARGONAUTE4 , one of the characteristic components of the RdDM pathway , was required for plant immunity against bacterial pathogens . Here , in a genetic screen aiming to identify cellular factors integral in regulating immunity in Arabidopsis , we further identified that the RNA polymerases V , another crucial component of the RdDM pathway , is pivotal for plant immunity against fungal pathogens . Similarly , we identified that additional components of the RdDM pathway , but surprisingly not RNA polymerase IV , are similarly required for plant immunity . Based on genetic evidence , our results highlight the importance of RdDM as an additional layer of complexity in the regulation of plant immune responses . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"plant",
"science",
"model",
"organisms",
"plant",
"biology",
"genetics",
"biology",
"genetics",
"and",
"genomics"
] | 2011 | The RNA Silencing Enzyme RNA Polymerase V Is Required for Plant Immunity |
Quantifying epidemiological dynamics is crucial for understanding and forecasting the spread of an epidemic . The coalescent and the birth-death model are used interchangeably to infer epidemiological parameters from the genealogical relationships of the pathogen population under study , which in turn are inferred from the pathogen genetic sequencing data . To compare the performance of these widely applied models , we performed a simulation study . We simulated phylogenetic trees under the constant rate birth-death model and the coalescent model with a deterministic exponentially growing infected population . For each tree , we re-estimated the epidemiological parameters using both a birth-death and a coalescent based method , implemented as an MCMC procedure in BEAST v2 . 0 . In our analyses that estimate the growth rate of an epidemic based on simulated birth-death trees , the point estimates such as the maximum a posteriori/maximum likelihood estimates are not very different . However , the estimates of uncertainty are very different . The birth-death model had a higher coverage than the coalescent model , i . e . contained the true value in the highest posterior density ( HPD ) interval more often ( 2–13% vs . 31–75% error ) . The coverage of the coalescent decreases with decreasing basic reproductive ratio and increasing sampling probability of infecteds . We hypothesize that the biases in the coalescent are due to the assumption of deterministic rather than stochastic population size changes . Both methods performed reasonably well when analyzing trees simulated under the coalescent . The methods can also identify other key epidemiological parameters as long as one of the parameters is fixed to its true value . In summary , when using genetic data to estimate epidemic dynamics , our results suggest that the birth-death method will be less sensitive to population fluctuations of early outbreaks than the coalescent method that assumes a deterministic exponentially growing infected population .
In many applications determining the past dynamics of populations is of interest . In an epidemiological context in particular , the interest lies in knowing two quantities: the basic reproductive ratio and the growth rate of the epidemic . is a key parameter that determines the probability and the extent of spread of the disease in the population . It is defined as the number of secondary infections a single pathogen is expected to cause when introduced into a population of susceptible individuals [1] , [2] . The growth rate determines the speed of spread of the pathogen . Accurate estimation of these two parameters ( and ) is required in order to take appropriate measures of intervention , e . g . vaccinations or isolation of infected individuals . Until recently , estimation of these parameters was exclusively based on prevalence and incidence epidemiological data . However , recent progress in phylodynamics has enabled the inference of these parameters from pathogen sequence data by integrating methods of phylogenetics with those of mathematical epidemiology ( for review see [3] ) . SIR-type models have been widely used to describe epidemiological dynamics [1] , [4] . In essence , these models are based on separating the population into different classes of individuals , namely susceptibles ( ) , infecteds ( ) and recovereds ( ) . Individuals can change their status , i . e . switch from one class to another . The epidemiological dynamics depend on two rates: a birth rate and death rate , where is the number of susceptibles and the total population size . The birth rate , or transmission rate , is the rate with which one infected individual will infect another uninfected individual . In the transmission tree , an infection event will be displayed as a bifurcation or split of one lineage into two lineages . The death rate , or removal rate , is the rate with which an infected individual becomes non-infectious , e . g . recovers from the disease , dies , or changes behavior . In the transmission tree , becoming non-infectious is a lineage that stops growing , i . e . becomes a tip in the tree . Various sampling schemes select a proportion of the infected individuals from the complete transmission chain to be included into the observed phylogeny . The observed phylogeny is the subtree induced by the complete transmission chain that connects the sampled individuals . This sub-selection of the individuals reflects the fact that in empirical datasets the pathogens of only a small fraction of the infected hosts have been sequenced and included into an epidemiological study . For an epidemic following SIR dynamics , the growth of the population size at the initial stage of the spread follows an exponential trend , although it slows down at later stages due to a depletion of susceptibles . We focus on a special scenario , where only the early epidemic outbreak , i . e . exponential growth of the infected population , is being considered . We can simplify the model to a constant rate birth-death process , where we assume no significant decrease in the number of susceptibles over time , formalized as , implying that birth rate is constant . Recent genetic sequencing efforts have produced many pathogen sequences from different hosts . To reconstruct their phylogenetic relationships , numerous methods have been developed ( refer to books [5] , [6] and references therein ) . The resulting phylogenetic trees are used as a proxy for the ( incomplete ) transmission tree , and thus provide us with insights into the dynamics of the epidemic . They enable us to estimate parameters such as transmission rate ( ) , removal rate ( ) , growth rate ( ) , or basic reproductive ratio ( ) . Methods based on Bayesian inference coupled with a Markov chain Monte Carlo ( MCMC ) procedure [7] infer the posterior distribution of trees ( ) together with the epidemiological parameters ( ) and sequence evolution parameters ( ) from genetic sequencing data based on the following relation: ( 1 ) Here , is the posterior distribution of the parameters and trees; is the likelihood of the parameters ( and ) that is usually computed by the Felsenstein algorithm [8]; is the probability density of the phylogeny given the epidemiological parameters ( e . g . assuming an SIR-type model ) ; and are priors for evolutionary and epidemiological parameters , respectively; and is the normalizing constant representing the integral of the numerator over all phylogenies and parameters . As data are fixed , is a constant and thus irrelevant for the estimation of the posterior probability density of the parameters in the MCMC procedure . Here , we focus on the impact of the underlying epidemiological model when calculating . Two models are mostly used in epidemiological contexts for this purpose: the coalescent [7] , [9]–[14] ( uses and review of the model are described in [15] ) and the birth-death process [16]–[25] ( reviewed in [26] ) . Both models have been used to estimate and/or the growth rate parameter of HCV [27] , [28] and HIV epidemics [24] , [25] , [29] . Since we only focus on phylogenies resulting from early epidemic outbreaks , we apply a special case of the birth-death model , namely the constant rate birth-death model with incomplete sampling , and a special case of the coalescent model , namely the coalescent with deterministic exponential infected population growth . Both models are implemented in the software package BEAST v2 . 0 [30] and have been used interchangeably for parameter inferences . The constant rate birth-death model implemented for parameter inference in BEAST v2 . 0 is precisely the epidemic outbreak model introduced above . The specific sampling scheme used in this study is the implementation of a constant sampling probability upon “death” ( happening with rate ) for each individual ( in BDSKY add-on of BEAST v2 . 0 ) , and is known as the incomplete sampling version of the birth-death model [24] . The coalescent with deterministic exponential infected population growth has been introduced in population genetics , and is now also used as an approximation for epidemiological dynamics . Classically , the coalescent has been used in phylodynamic studies . The coalescent reconstructs the ancestry of sampled individuals towards the most recent common ancestor ( MRCA ) . In fact , it reconstructs the probabilistic structure of the tree by merging lineages progressively going back in time as a function of the population size until there is only a single lineage left [31] . The coalescent thus provides a prior distribution of trees given a population size , where the population size may change through time . In the epidemiological context the population size of interest is that of the infected individuals . This probability density function allows for the calculation of the probability of the tree for given population size parameters [11] , [12] , [29] . The coalescent seems to be a good approximation to many processes arising in biology ( e . g . [11] , [28] , [32] ) . However , violations of the model assumptions can lead to consequences whose nature and extent are still not fully explored [24] , [29] . The coalescent can be interpreted as a continuous time approximation of the discrete time Wright-Fisher model [33] , [34] . Based on this approximation , as stated by Rodrigo and Felsenstein in [29] , the requirements for the studied population to be well approximated by the coalescent are: In most of the coalescent models , a further assumption is made: The coalescent can also be interpreted as a continuous time limit of the discrete time Moran process [35] . The assumptions above with exception of 1 ) are also required for the continuous time approximations of the discrete time Moran model , rather than Wright-Fisher population model . Continuous time versions of Wright-Fisher and Moran population models can also be formulated directly rather than by approximation of discrete time models . This is done by assuming a rate of coalescence in continuous time instead of approximating it by a conversion from discrete to continuous time space , as we point out in the Supplementary Material S1 . Such continuous time Wright-Fisher and Moran population models can be formulated as a coalescent process without the assumptions 1 ) and 3 ) . Furthermore , extensions to avoid deterministic population size changes , i . e . assumption 5 ) , have been developed [36] . In an epidemiological context , the sampling proportion in a recent epidemic can be quite high . For instance , the sampling proportion of the HIV epidemic in Switzerland has been estimated to be 0 . 75 [37] . This high sampling proportion is a misspecification for the discrete time Wright-Fisher and Moran process-based coalescent models . Although some studies suggest that the violation of this assumption should not be of significant importance [38] , its consequences on the model performance when applied to empirical data are so far unknown . In addition , impacts on the parameter inference under the coalescent in the context of deviations from the deterministic population size assumption are not well explored either . The birth-death model also requires assumptions 2 ) and 4 ) to be fulfilled . Additionally , the generation times are assumed to be exponentially distributed , instead of discrete generations as in the assumption 1 ) . We see the major difference between the birth-death model and the coalescent in three factors: In this paper , we want to shed light on the practical importance of the theoretical points raised above for parameter estimation . We investigate the comparative performance of the birth-death and the deterministic coalescent model in phylodynamic parameter estimation by doing a simulation study . We first simulated both constant rate birth-death model trees with incomplete sampling ( from now on simply referred to as the birth-death model , unless specified otherwise ) , and coalescent model trees with deterministic exponential infected population growth ( from now on simply referred to as the coalescent model , unless specified otherwise ) . We then applied phylodynamic methods based on the birth-death model and the coalescent model within the BEAST v2 . 0 software package to the simulated phylogenies . In this fashion , we estimate the phylodynamic parameters and compare coverage ( measured as the fraction of simulated trees where the HPD captures the true parameter ) , accuracy ( measured as the root mean square error ( RMSE ) ) and precision ( measured as the width of HPD intervals ) of the parameter estimates .
The ability of this specific birth-death model and the coalescent model to capture the true growth rate in the 95% highest posterior density ( HPD ) interval at fixed is summarized in Table 1 , and shown in Figure 1 , Figure S1 and Figure S2 . Note that none of these , nor any of the results below , change substantially if we use quantiles instead of HPD intervals ( data not shown ) . For , the birth-death model successfully recovers the growth rate parameter for trees simulated under the birth-death model , whereas the coalescent model successfully recovers the growth rate parameter for those trees that were simulated under the coalescent . This observation is not surprising but confirms the basic expectations that the model used for simulation should be good when it is also applied for inference . In the critical case where , the birth-death model recovers the true growth rate only in 78% of the birth-death trees . This is because the birth-death likelihood is conditioned on the time of origin of the process ( length of epidemic ) . Our simulated trees are all of different lengths , however , as we stop once reaching 100 tips . That means that for low growth rates , we select a very biased set of relatively big trees , as most realizations would die out before producing 100 tips . By looking at Figure S1 we observed that especially for low values , i . e . and , the selective inclusion of the relatively big trees into the final set results in the median estimates of the growth rate parameter to be biased towards higher values than the truth . To show that the birth-death inference method has no bias if applied to trees with a large fixed time of origin , we simulated trees under , , until reaching fixed time , or ( Figure S3 ) . The 95% HPD intervals of the birth-death model growth rate estimate capture the true value of the growth rate in 95% and 96% of the cases , respectively . The distribution of the medians of the 100 HPD intervals is spread out evenly around the true value of , meaning the effect of growth rate over-estimation decreases for increasing . When applying the birth-death method to coalescent trees , the growth rate coverage is higher than when applying the coalescent method to birth-death trees . The higher coverage of the birth-death model comes partially at the cost of a larger 95% HPD interval size ( see Table 1 ) . The normalized 95% HPD interval sizes of produced by the birth-death process and by the coalescent are almost identical for very large ( ) , and they increase for both coalescent and birth-death model parameter estimates with decreasing . However , the coalescent intervals become smaller than birth-death intervals , and at their widths differ by a factor . This discrepancy between the HPD sizes does not translate in decreased accuracy of the birth-death model compared to the coalescent model . Accuracy can be measured by the root mean square error ( RMSE ) of the median of each posterior interval . The accuracy of the birth-death model is higher ( RMSE is lower ) than the accuracy of the coalescent when applied to the birth-death trees . The accuracy of the birth-death model is lower than the accuracy of the coalescent model on the coalescent trees for our analyses with . As expected , the above observations do not change when the branch lengths , i . e the time units , are scaled . This corresponds to multiplying the birth and the death rates , and thus the growth rate parameter , by a constant factor while keeping unchanged ( Table S1 ) . The increase in HPD interval size when lowering may be due to increasing stochasticity in population size variation over time . This increased stochasticity is caused by decreased population growth resulting from more death events per birth event . We will now discuss the reason for biases when applying the birth-death method to coalescent trees and vice versa . When applying the birth-death method to coalescent trees , for , the true growth rate is recovered very reliably . The birth-death process has a small bias when applied to the coalescent trees for and . This can be explained again by the simulation scheme and vanishes if simulating for a fixed time rather than until a number of samples is reached ( Figure S3 ) . When applying the coalescent method to birth-death trees , the coalescent misses the true growth rate value ( the HPD does not contain the true growth rate ) on the birth-death model trees more often than the birth-death model on the coalescent trees . This is accentuated with decreasing . The coalescent has a tendency to overestimate the growth rate ( see Figure 1 ( ) and Figure S1 ) . The growth rate overestimation can be explained by the push-of-the-past effect described by Nee et al [39] . The exponential growth coalescent model assumes a constant population growth rate . The push-of-the-past effect causes the expected population size under the birth-death model to initially increase faster than with rate , and then slow down to grow with the constant rate . As a consequence , when a final population size is fixed the coalescent trees are predicted to be longer than the birth-death trees . Put in other words , given a fixed time after the start of the process , the expected birth-death population size is bigger than the expected coalescent population size . This push-of-the-past effect becomes more severe for smaller values of . For inference , this means that the coalescent applied to birth-death trees infers inflated growth rates as coalescent trees with the true growth rate would be expected to be longer . We observe this push-of-the-past effect and the resulting differences in coalescent and birth-death trees in our simulations . We investigated the difference between birth-death and coalescent trees for the parameter combinations where most stochasticity in population size variation over time and most bias of the inferred parameters was observed: , i . e . . Both the mean and the median measure of the tree lengths without the root-origin distance showed that the coalescent had a strong preference to produce overall longer simulated trees . Median tree length of the coalescent trees was 35 . 2 while that of birth-death trees was only 26 . 1 . Similarly , the mean tree length was 36 . 1 and 27 . 7 for coalescent and birth-death trees , respectively . Upon visual inspection of these trees , we noticed that the coalescent mostly produced trees with longer inner branches , especially close to the root , as compared to trees simulated under the birth-death model . As the population size growth curves in semi-logarithmic plots are parallel lines with the slope for large , the relative difference between the coalescent and birth-death tree lengths should become smaller for increasing . This means that the ratio of the length of the birth-death tree and of the length of the coalescent tree tends to 1 when letting the trees ( populations ) grow for longer times . In fact , by increasing the final population size through sampling more tips ( , 500 or 1000 ) and thereby suppressing the importance of the push-of-the-past , the median of the coalescent estimates of the growth rate gets closer to the true value . Nevertheless , the coverage of the coalescent model does not improve due to overconfident estimates , i . e . shrinking HPD intervals ( Figure S4 ) . When applying the birth-death model to the long coalescent trees we occasionally observed an overestimation of the growth rate . In most cases the growth rate is estimated correctly by the birth-death model though . The coalescent trees typically have long branches close to the root . In the following we will demonstrate that these early long branches do not strongly impact the overall likelihood calculation in the birth-death model . To investigate the impact of long early branches on parameter estimation under each model further , we changed the simulated trees systematically . We looked at trees simulated under and . In these trees , we extended each branch existing after 10% , 20% , 30% , 40% , 50% , 60% , 70% , 80% , 90% of the time between the root an the present . We extended branches in all these cases by 48 for and by 0 . 18 for , which was approximately the full length of the birth-death tree ( Figure 2 and Figure S5 ) . Re-analyzing the trees with using the birth-death model and the coalescent model revealed high sensitivity of the coalescent estimates to early but not so much to the late perturbations . The birth-death model estimates of the growth rate proved to be more robust to early perturbations than to late perturbations , see Figure 2 . We note that the early perturbations actually only affect very few branches , and thus only introduce minor changes to the data . We hypothesize that the ability of the birth-death model to account for stochasticity in population size determines its robustness to above-introduced changes in branching times . When perturbing the tree close to present , the birth-death model growth rate estimates decrease as all branches occurring prior to the perturbation are not allowed to produce sampled descendants later on . For that particular reason , the birth-death method infers a much too low median growth rate when we only extend the single branch leading to the most recent bifurcation ( data not shown ) . Thus , the birth-death method is very sensitive to even minor perturbations close to the present . Re-analyzing the trees with revealed that both the birth-death and the coalescent method are very sensitive to perturbations ( Figure S5 ) . This is most likely because the process is very deterministic at high ( Figure S2 ) . The perturbation cannot be considered to be the result of stochastic population size changes as at low , and therefore they significantly influence the overall likelihood values of the inference methods . Another way to simulate trees under a different model than the birth-death or the coalescent model is to simulate SIS/SIR trees . In both these alternative models , there is an exponential growth of the infected population early on , followed by a decrease in transmissions ( births ) due to a depletion of susceptibles . In trees produced by an SIS model with small total population size ( ) , the curve of infected individuals over time follows a logistic trend . In the case of the SIR model assuming a small total population size the growth slows down after the exponential phase , then stops and finally becomes negative , i . e . the population size of infected individuals decreases . When applying the exponential growth birth-death and coalescent models to the SIS/SIR trees , the birth-death process underestimates the growth rate more severely than the coalescent model in trees which reach the post-exponential growth phase . This fits very well to the results presented above for the constant rate birth-death trees . The birth-death model uses all available information in the tree , and thus obtains an average growth rate estimate from the SIS/SIR trees which is lower than the initial growth rate due to the post-exponential slowdown in transmission . The coalescent mainly uses information from the early epidemic . It thus puts less emphasis on the post-exponential phase and consequently does not severely underestimate the growth rate . The birth-death model does not produce consistently larger HPD intervals than the coalescent ( see Table 2 ) , in contrast to the exponential growth trees . In summary , the coalescent model estimates of the growth rate seem to be influenced most strongly by the early branching patterns in the tree . These early patterns most strongly reflect stochasticity in population size . In contrast , the birth-death method averages the information throughout the tree . Since the sampling probability is fixed to quite a high value ( ) in all the trees simulated above , the trees are not only relatively short , but also a lower translates to more death events per birth event , and consequently means higher sampling from the population ( as we sample from the individuals that become non-infectious with sampling probability ) . We further investigated if this relatively high sampling causes the coalescent methods to fail for low . We simulated trees at low , medium and high , , but this time also at different sampling probabilities ( see Table 3 ) . For some parameter settings with very low sampling probability , the tree simulations did not finish within 7 days of simulation , and the results are thus not displayed in the summary table . We observed that the size of the 95% HPD interval become smaller with lower sampling probability for both birth-death and coalescent estimates of the growth rate . This means that both methods become increasingly confident in the growth rate estimates with increasing tree length due to decreased sampling . Additionally , the smaller the sampling probability in the birth-death tree simulation , the more often the true growth rate parameter is recovered by the coalescent . The same is observed for the growth rate parameter estimate produced by the birth-death model on the coalescent trees . In fact there are two ways to grow the tree longer , either by decreasing the sampling probability or by sampling more tips . As discussed in the previous section , growing the tree longer decreases the push-of-the-past effect . In contrast to decreasing the sampling probability , the coverage of the coalescent method does not improve when sampling more tips . This is presumably because stochastic effects are not diluted when sampling densely . This is best seen when comparing the coverage of the coalescent on birth-death trees grown for longer by increasing the final sample size ( as in Figure S4 ) to that on the trees grown longer by decreasing the sampling probability ( Table 3 ) . An especially informative comparison at can be made between trees where the final sample is equal to 10000 , the average tree length of which ( root-origin distance not included ) is 105 . 9 , and the coverage is 8/100 ( figure not shown ) and those trees grown with , reaching average length of 99 . 7 and coverage of 55/100 . Overall , the coalescent struggles most with correct growth rate estimation for datasets with low and high sampling probability . At low , compared to high , we have a strong push-of-the-past effect and remain for longer in the phase of strong stochastic changes in population size over time . A high sampling probability means that most samples are taken in the early phases of the epidemic , the phase with the push-of-the-past effect and reflecting stochasticity in population size the most . It could be argued that a high sampling probability which leads to a high sampling proportion is a violation of one of the main assumptions of the coalescent model and the main reason for the biases of the coalescent when applied to birth-death trees . Indeed , for the discrete time Wright-Fisher and Moran model , we have to assume a small sampling proportion when deriving the continuous time coalescent approximation . However , as we show in the Supplementary Material S1 , the coalescent can also be interpreted as a continuous time Wright-Fisher and Moran model , and these models do not require a small sampling proportion . In fact , one can even assume complete sampling , i . e . . Therefore , we suspect that the high sampling proportion just unmasks the real reason for the frequent inability of the coalescent method to include the true value of the growth rate parameter of the birth-death trees in the 95% HPD interval . The real reason being the stochastic population size variation over time . Finally , we investigated the sensitivity of the models towards variation in sampling schemes . We simulated trees where periods of no , ( or low , ) , sampling at the beginning were followed by a period of complete sampling , . Furthermore , we simulated trees where a period of initial complete sampling was followed by a period of no ( or low ) sampling , and later again followed by a period of complete sampling ( Figure 3 and Figure S6 ) . The coalescent model is very robust to these changes in sampling schemes . The birth-death model is robust to slight sampling variations , but overestimates growth rate severely for extreme changes to the sampling scheme in particular for high . Use of the birth-death skyline model , assuming a time-varying sampling probability rather than a constant sampling probability , reduces this bias ( Figure 3 and Figure S6 , Table S2 ) . So far we only investigated the inference of epidemic growth rate using the birth-death and the coalescent models . Both models also estimate other epidemiological parameters . The birth-death model is parameterized by the transmission rate , becoming-non-infectious rate and the sampling probability . The coalescent model is parameterized by , and . These parameters as well as the compound parameter can be inferred given we fix one of the three model-specific parameters . For the birth-death process , so far we fixed to the true value during the analyses . Now we investigate to what extent we can estimate the individual parameters , including , using the birth-death method . We re-analyzed all birth-death and coalescent trees simulated above applying the birth-death model estimating , and setting to , , ( and/or , if this was used for tree simulation ) , and not fixed sampling probability but assume a uniform prior over interval . For example , trees produced under were analyzed under ( true ) , , and . The likelihood of a tree only depends on and , rather than on three parameters , , [25] . We could confirm that no matter what is used for the analysis , true growth rate is equally well estimated by the birth-death process for both stochastic birth-death trees or coalescent trees ( Figure S7 shows results for trees simulated under ) . The same holds for estimation of ( Figure S8 displays the results for trees simulated under ) . During this analysis , we also noticed when we set to its true value ( i . e . the value used during the tree simulation ) , we are able to recover the true and parameters , and consequently also the true from both the trees generated under the birth-death model and those generated under the coalescent model ( see Figures S9 , S10 , S11 and S12 ) . In the Supplementary Material S1 , Section “Parameter correlations under the birth-death process” , we show analytically that for fixed and , increases , and and decrease with increasing , and vice versa . In Figure 4 , we plot the impact of changing on the value . We confirmed this theoretically predicted bias in parameter estimation in our simulation study . If we fixed during the birth-death analysis to a bigger value than the true used during the birth-death simulations , then we overestimated and underestimated and , and vice versa . Similarly , when analyzing the coalescent trees with the birth-death model , we observed an upward shift for and downward shift for and when assuming a value bigger than used in the simulation of the sampling times , and vice versa . Using a uniform prior for over the interval had different effects on estimation of , and , depending on the sampling probability used for simulation . First , in cases where the true , use of uniform prior for during the analysis resulted in wider 95% HPD intervals that either fully , or mostly , contained the 95% HPD interval produced when was fixed to the true value ( Figures S7 , S8 , S9 , S10 , S11 and S12 ) . This is because the value is the median of the prior . Second , for simulated trees with a true , the 95% HPD intervals produced using a uniform prior on were shifted away from the 95% HPD intervals that resulted from analysis where was fixed to the true value ( data not shown ) . As predicted by derivations in the Supplementary Material S1 , for a true below , the estimated interval for and was shifted downward , compared to the interval estimated when the was fixed to the true value , and the estimated interval for was shifted upward . When the true used for simulations was higher than , the posterior intervals for and shifted upwards , whereas the posterior interval for shifted downwards . Overall , the birth-death method recovers two out of the three individual epidemiological parameters reliably if one of these parameters is fixed ( here ) . The epidemic growth rate can be recovered well even if is misspecified . If , or any of the other two parameters , or , is set to the true value , we can recover ( Figures 5 , S13 and S14 , for fixed ) . If any of the parameters is fixed to a wrong value , e . g . if one assumes incorrect , then the original ( true ) cannot be recovered . Equivalently , when using the coalescent for inference , and knowing one of the parameters , or present day infected population size , we can also recover the parameter , given we estimated the growth rate correctly ( Figure 5 and Figure S14 display the scenario where is known ) . We use the transformation [40] to obtain estimates from the posterior estimates of the growth rates . The birth-death inference method is partially informed by the sampling times , as the sampling times are outcomes of the birth-death process with constant rate sampling . The coalescent is only informed by the branching times in the tree , as the coalescent conditions on sampling . In order to compare the ability of the coalescent and the birth-death model to infer the underlying population dynamic parameters in the absence of information on sequentially sampled tips , we simulated 100 trees on 100 tips sampled at one point in time under the birth-death model . The simulation was done with ‘TreeSim v2 . 0’ on CRAN [41] ( function sim . bd . taxa ) . We assumed the two extremes , and with . We assumed a uniform prior on the length of the epidemic and sampled each infected individual at time with sampling probability . For parameter inference , we again used the birth-death model in BEAST v2 . 0 . As the choice of during inference does not influence the growth rate estimate [22] , analog to fixing in the sequential sampling model not influencing the growth rate estimate , we fixed to the truth . For the birth-death model , the true growth rate was contained in the 95% HPD intervals in 91–96 out of 100 cases . For the coalescent , the true growth rate was contained in the 95% HPD interval in 63–75 out of 100 cases . The increased coverage of the coalescent model compared to the sequential model is likely due to two factors: 1 ) the trees are in general longer meaning the early stochastic fluctuations are less dominant , and 2 ) there is no sampled data during the early very stochastic changes in population size , as all samples are collected only at the end of the epidemic . The coalescent still does not reach more than 80% coverage because similar to the case of sequential sampling , too narrow HPD intervals are being inferred , see Table S3 , and Figures 6 , S15 and S16 . Overall , when using phylogenies generated under the birth-death process , the birth-death model produces better results than the coalescent even in the absence of sequential sampling time information . Thus , when analyzing sequentially sampled phylogenies , the better performance of the birth-death model does not exclusively come from using the information of the tip times explicitly . This supports our claim that the main reason for the birth-death model in general producing more reliable estimates comes from it taking stochastic population size changes explicitly into account . As seen above , the coalescent is often overconfident , i . e . it estimates too narrow 95% HPD intervals , and thus often does not contain the true growth rate parameter . For comparison , we calculated point estimates both for trees with tips sampled sequentially and for trees with tips sampled at one point in time . For various parameter combinations we determined the maximum a posteriori ( MAP ) estimate from the final MCMC runs . We also report the maximum of the tree likelihood values ( ML ) of the final MCMC run . For uniform priors and fixed time to origin , the ML corresponds to the maximum likelihood estimate . We verified this for trees with sequentially sampled tips by estimating the maximum likelihood values with ‘TreePar v3 . 0’ on CRAN [42] . In contrast to the findings obtained when using the HPD intervals , Figures 7 , S17 , S18 and S16 show that the 100 ML and 100 MAP point estimates on the birth-death trees do not differ much for the coalescent and the birth-death inference method . Similarly to the results obtained when using HPD intervals for the coalescent trees , the birth-death model overestimates growth rates for high and and . This highlights that when evaluating the performance of different inference methods , we should always consider the HPD/credible/confidence intervals in addition to the point estimates .
Under both the birth-death and the coalescent models , the times of coalescence or bifurcation are stochastically selected from the distribution of coalescence and bifurcation event times . The individual trees produced by both the coalescent and the birth-death process are thus stochastic realizations of the respective processes . However , the coalescent model with exponential growth of the infected population assumes deterministic changes in the population size . Therefore , the trajectory of the infected population size follows an exponential growth . It has been pointed out before [31] that the coalescent model can appropriately approximate population dynamics reflecting models where the sampled genealogy is conditioned on the total population size that varies deterministically . It has been postulated earlier [29] that coalescent approximations are good for datasets where the sample size is sufficiently smaller than the population size . On one hand , this approximation is necessary when translating the discrete time Wright-Fisher and Moran model population dynamics into a continuous time coalescent framework . In fact , such an inequality should hold also when is the number of co-existing lineages in the tree at any time point in the process . On the other hand , the assumption of small sampling proportion ( ) is not required when the coalescent is interpreted as a continuous time Wright-Fisher and Moran model , and has been found unimportant for some inferences based on the Kingman coalescent [38] . We show that the ability of the coalescent with deterministic exponentially growing infected population to properly estimate the dynamical parameters of the birth-death trees derived from early epidemics is questionable , which was previously suggested by the small simulation study in [24] . It often struggles with , and overestimates the growth rate parameter from the birth-death trees representing samples from early epidemics starting with one single infected individual . This effect is accentuated when decreasing the basic reproductive ratio and increasing the sampling probability . These settings ( small and high ) produce birth-death trees that display a significant push-of-the-past effect , and thus differ in their length ( age ) the most from the coalescent model trees . Furthermore , both increasing the sampling probability and decreasing lead to smaller infected population sizes while samples are collected , which in turn corresponds to more stochasticity in population size changes over time reflected in the final trees . Thus , we hypothesize that the coalescent model with deterministic exponentially growing population fails to recover population dynamic parameters when stochasticity is important . This claim is supported by three observations . First , the birth-death trees show systematic difference from the coalescent trees in their lengths due to the push-of-the-past effect . This effect is the strongest in low regimes , the regime where we observe that the coalescent struggles most with parameter recovery . Second , early stochastic effects in the population size changes are not well captured by the coalescent . This became apparent when we introduced artificial perturbations to branching times early in the tree , which had a large effect on parameter estimates . In fact , the early phase in the tree informs the coalescent parameters the most , and late perturbations in the tree have little impact on parameter estimates . Additionally , increasing the sampling probability , especially at medium and low , where the population growth in the constant rate birth-death model displays much stochasticity , results in the coalescent failing most often . Even if we increase the final infected population size by sampling more tips , i . e . sampling for a longer time period , this effect is not corrected for and the coalescent has a very low coverage . Third , the coalescent does not capture the true growth rates in many cases because the HPD intervals are narrow around the median estimate . Short HPD intervals likely result from the coalescent model only considering one population size trajectory , namely the deterministically growing exponential population size , per epidemiological parameter combination and ignoring any stochastic uncertainty . The obvious solution to these problems would be to use the coalescent that accounts for stochastic population sizes as in Rasmussen et al . [36] . It is expected that such a model would show an increased coverage rate , but also increased HPD interval sizes , similar to those produced by the birth-death model . It would also be interesting to investigate the performance of the two models on the trees resulting from the epidemics starting with more than a single individual . In contrast to the coalescent , stochasticity in population size is incorporated into the birth-death process . The birth-death model performs well not only on birth-death trees but also on most coalescent trees . This is due to the birth-death process interpreting variation in branching times as stochastic changes in population size over time leading to large HPD intervals . The birth-death model therefore accepts many different parameter combinations that may have given rise to the observed tree . Furthermore , the ability of the birth-death model to correctly estimate the parameters from the coalescent trees decreased only in those trees that were simulated under combination of high and low . However , unlike in the case of the deterministic coalescent model , in this case the bias can be easily corrected for by letting the trees grow for longer , by sampling more tips or by decreasing the sampling probability . It could be argued that the superior performance of the birth-death model in terms of coverage of the true growth rate comes from explicit usage of the sequential tip sampling times . As we have shown here , even in the absence of information from the sequentially sampled tip times , the birth-death model shows higher coverage than the deterministic coalescent model . This means that it is not only the sampling process that improves the coverage of the birth-death model , but the branching time information alone already leads to better estimates under the birth-death model compared to the deterministic coalescent . The birth-death model is sensitive if large parts of the tree depart from the birth-death model assumption , as the birth-death model averages over observations throughout the whole tree . This sensitivity becomes particularly apparent in the analysis of SIS/SIR trees and of trees with varying sampling probability over time . Both the birth-death skyline [25] and coalescent skyline plot [28] method aim to capture transmission and removal rate changes over time . Furthermore , the birth-death skyline plot accounts for the sampling probability to change throughout the tree . If the sampling probability changes frequently , it might become hard to obtain good estimates for it . The coalescent conditions on the sampling times and hence does not face the problem of estimating highly varying sampling probabilities . It however still assumes that at the time of sampling , a lineage is chosen uniformly at random from all co-existing lineages . Developing a birth-death based inference tool conditioning on sampling might make use of the advantages of the birth-death and the coalescent tools: accurate inference of HPD intervals by the birth-death method and robustness to time-heterogeneous sampling by the coalescent . In case of lineage-specific sampling , multi-type birth-death models can be employed [43] . The differences of the birth-death method and the coalescent method are mainly observed in the 95% HPD intervals , while only minor differences were apparent in the point estimates , i . e . the maximum a posteriori and maximum likelihood estimates . When comparing the performance of inference methods , it is therefore crucial to always assess the capability of recovering the uncertainty in parameter estimates . While our study only compared the birth-death and the coalescent methods on simulated trees , previous work also compared the two models on empirical data . A higher estimate of the growth rate parameter of the HCV epidemic in Egypt was reported previously when data were analyzed with the coalescent model as compared to the birth-death model [22] . Furthermore , the coalescent posterior for growth rate displayed a larger spread ( standard deviation ) compared to the birth-death model . The larger standard deviation should indicate larger HPD intervals . Similarly , in [24] , a higher growth rate estimate with larger 95% HPD interval size was observed for Swiss HIV under the coalescent compared to the birth-death model . As for our birth-death simulated trees , the HPD intervals obtained using the coalescent model compared to the birth-death model are shorter , the results based on empirical data contradict these simulation results . However , results from both of these studies are consistent with the epidemic following either the SIS or the SIR-type dynamic , already reaching the post-exponential growth phase ( compare with our simulation results in Table 2 ) . Indeed , application of the SIS model to the HIV epidemic in Switzerland rejects the simple birth-death model and reveals a higher [44] than previously estimated [24] . In the present study we estimate epidemiological parameters based on fixed trees . We note that the tree probability only depends on the branching ( transmission ) and removal times . Thus , the Bayesian estimates of growth rate from trees under full sampling , i . e . , are equivalent to estimating the posterior growth rate from full incidence and prevalence data . Birth-death based methods explicitly parameterize epidemiological parameters beyond the growth rate . Previously reported correlations when estimating , , and simultaneously [22] , [25] are confirmed by our simulation results . In fact , we can only estimate two compound parameters , namely growth rate and , when we have no information beyond the phylogeny . If we are only interested in these two parameters , we can fix one of the tree individual parameters , e . g . sampling probability , to an arbitrary value . Caution must be taken , however , when drawing any conclusions on , , and . Fixing one of these parameters during the data analysis to a wrong value will produce biased estimates of the remaining parameters . Similarly , , , and can be estimated by the coalescent model , given that one of these parameters is known , and conditioned on the fact that the growth rate and scaled population size were inferred correctly . In general , when one knows what kind of population dynamic process gave rise to the tree , the appropriate method should be applied for phylodynamic parameter estimation based on such trees . In case of doubt of the underlying process generating the genealogy , our results show that in scenarios of constant sampling and exponential population growth , especially when the samples were drawn early in the epidemic outbreak , the constant rate birth-death process with incomplete sampling is a better choice than the coalescent assuming deterministic exponential infected population growth . The assumption of exponential spread of a pathogen with constant sampling is limiting since many epidemics are better characterized by SIS or SIR dynamics with time-varying sampling , for example . Recent work proposed birth-death based SIS and SIR models [44] , [45] with time-varying sampling [25] , and coalescent based SIS and SIR models [13] , [14] for phylodynamic inference . It is not clear how important stochasticity is for the epidemiological SIS/SIR models . Future work should thus focus on a detailed comparison of birth-death based and coalescent based SIS and SIR models , such that the empirical data can be analyzed and interpreted using appropriate methods .
We used three epidemiological SIR-type models to simulate transmission trees growing forward in time , namely the constant rate birth-death process ( epidemic outbreak ) , the SIS model and the SIR model . All of the models are implemented as a Gillespie algorithm [46] in the R package ‘TreeSim v2 . 0’ on CRAN [41] in the function sim . bdsky . stt . All models have a common birth or transmission rate with which one infected individual transmits the pathogen , where is the total population size assumed to be constant . In the birth-death model , the impact of the susceptible population size is assumed to be neglected and thus meaning transmission rate is constant . In all models , infected individuals become non-infectious with rate . In the SIS model , a recovered ( removed/non-infectious ) individual goes back into the susceptible class , while in the SIR model , a recovered individual goes into the recovered class . Note that in the constant rate birth-death model , the fate of a recovered individual does not have to be modelled , as the number of susceptibles does not influence transmission rate . These models induce transmission trees in the following way . Each infected individual is represented by a lineage . A transmission event results in a branching event , while a removal event results in the termination of a lineage , thus a tip in the tree . We assume that we sample each tip from the complete transmission tree with probability , acknowledging that in empirical data only a fraction of the infected individuals are sampled and thus included into the dataset . The tree on the sampled tips is called the ( observed ) phylogeny , on which we do all our analyses . Note that the sampled tips are spaced through time , i . e . serially sampled . Initially , the population size of infected individuals simulated by these models is increasing at the rate . Thus , in expectation , the total infected population size at time before present time is . The constant rate birth-death process always stays in the exponential growth phase . Under the SIS model , we have a finite total population size , resulting in expectation in a logistic curve of number of infecteds over time: the initial exponential growth phase is followed by the slow down ( saturation ) of the apparent growth rate of the epidemic until the equilibrium is reached and the growth ceases . Under the SIR model , there is only a single one-way flow of individuals from to to and those that recover cannot become susceptible again , nor are they replaced by new susceptible individuals . The total population size is again limited and constant , . Given the definition of the model , the number of infecteds over time first increases exponentially , then slows down ( when there are still susceptible individuals available ) reaching a peak , and then declines to . The number of susceptibles in turn constantly decreases , and the number of recovered individuals constantly increases . For each chosen parameter setting of , we simulate 100 trees . We stop each tree simulation once we reach 100 serially sampled tips . We produce trees generated both by SIR-type models ( constant rate birth-death , SIS and SIR models above ) and by coalescent models for optimal comparison of performance of the birth-death and the coalescent model in estimation of phylodynamic parameters . We used the BEAST v2 . 0 package [30] to simulate coalescent trees with temporarily spaced leaves by sampling from the prior distribution of trees generated by the coalescent model with deterministic exponentially growing population ( example xml is in File S1 ) . For the coalescent simulation , we used the same parameter settings as in the birth-death simulations whenever possible . Thus , we used and to specify the rate of exponential population growth ( as under the birth-death model the population size also grows exponentially in expectation ) , as well as generation time ( inter-transmission interval length = [14] ) . As the coalescent does not model a sampling process , we conditioned the sampling times for each of the 100 coalescent simulations to those obtained by the birth-death simulation . Further , we assumed the present day infected population size to match the final infected population size in the birth-death simulation . During each coalescent tree simulation , we sampled 10 , 000 , 000 trees from the prior and chose the last one ( 10 , 000 , 000th ) to be analyzed by the birth-death model and the coalescent model in order to infer the growth rate parameter . The constant rate birth-death model with incomplete sampling and the coalescent model with deterministic exponential population growth were applied to the simulated trees to infer the posterior distribution of parameters . For this purpose , we used equation ( 1 ) . As we did all analyses on fixed trees and did not use sequence data , only and are relevant , all other terms are constant . For the analysis of the simulated trees , we assumed that the epidemiological parameters did not change at any time during the time span encompassed by the phylogenetic tree; meaning we assumed simple exponential growth of the epidemic . We explored performance of the MCMC implementation of the birth-death skyline serial model with 1 interval and the coalescent model with exponential growth rate as implemented in BEAST v2 . 0 ( birth-death skyline model in add-on BDSKY [25] ) . For the analyses explicitly mentioning use of the birth-death skyline model , we used 10 intervals for the sampling probability . The expression for under the exponential growth coalescent and the birth-death model have been derived previously [12] , [23] , [24] . As the goal of our paper is to identify the impact of using either of these formulations for inferring the epidemiological parameters , we will state the mathematical expressions here . For stating under the coalescent , we need the following definitions . We measure time going backwards from present . Present time is defined to be 0 . Let be the growth rate , be the duration of one generation in calendar units ( equal to the inter-transmission interval length ) [14] , and , which leads to . Thus , is the instantaneous rate of any pair of lineages merging . Let be an internal node at time of the tree ( if we have leaves in the tree , the number of internal nodes is ) , corresponding to , going back in time , a coalescent event . Let be a tip at time ( total of ) and be the number of lineages co-existing in the interval between time and , where is the time of the node ( internal node or tip ) occurring directly after ( i . e . more recent ) than node ( if = 0 , we set = 0 ) . Then , from [12] , corrected by [44] , ( 2 ) We assume throughout that the effective population size equals the infected population size , i . e . . For calculating under the birth-death model with serial sampling , we need the following definitions: Note that is the probability that an individual at time in the past will have no sampled descendants [23] . The probability under the birth-death process , conditioned on the epidemic circulating for time before the present ( i . e . the first lineage appeared at time ) , and conditioned on sampling at least one infected individual , is [23] , [24] , ( 3 ) For sampling all tips at one point in time , we again use the birth-death skyline model in add-on BDSKY [25] , now with sampling probability through time being 0 and present-day sampling probability being . To sample from the posterior distribution of the parameters of interest , we apply the Markov chain Monte Carlo ( MCMC ) computational procedure , which explores the posterior parameter surface by taking samples from their combined posterior distribution . We run the MCMC chain in BEAST v2 . 0 [30] for 2 , 000 , 000 steps and sample every 1 , 000th step to obtain an effective sampling size ( ESS ) for each parameter of 800 , or higher . For this purpose , we examined the log output file from BEAST in Tracer [47] . We remove the first 200 , 000 steps ( 10% ) as burn-in . For each parameter on each simulated tree , we plot the 95% highest posterior density ( HPD ) interval , meaning the shortest interval containing 95% of the posterior samples . We picked uniform priors for all parameters . Priors for the two parameters in the coalescent model with deterministic exponential population growth ( available in BEAST v2 . 0 ) were chosen as: In the birth-death model , we put priors on . We used the birth-death skyline serial model , setting the interval number to 1 , available in the BEAST v2 . 0 add-on BDSKY [25]: For the birth-death skyline analyses we set the interval number for sampling probability to 10 . For sampling all the tips at one point in time , is fixed to 0 and is set to the value used for tree simulation . For the remaining parameters , the same priors as above were used . | The control or prediction of an epidemic outbreak requires the quantification of the parameters of transmission and recovery . These parameters can be inferred from phylogenetic relationships among the pathogen strains isolated from infected individuals . The coalescent and the birth-death process are two mathematical models commonly used in such inferences . No benchmark on the performance of these models currently exists . We aimed to objectively compare two specific models , namely the constant rate birth-death model and the coalescent with a deterministic exponentially growing infected population . We compare coverage , accuracy , and precision with which they can capture the true epidemic growth rate parameter using simulated datasets . We find that the constant rate birth-death process can account for early stochasticity and is thus capable of recovering the epidemic growth rates more successfully . Provided one of the parameters is known , e . g . the sampling proportion of infected individuals , then the basic reproductive ratio can also be estimated reliably . We conclude that a birth-death-based method is generally a more reliable method than a deterministic coalescent-based method for epidemiological parameter inference from phylogenies representing epidemic outbreaks . Care should be taken if sampling is not constant through time or across individuals , such scenarios require so-called birth-death skyline models or multi-type birth-death models . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"phylogenetics",
"plant",
"science",
"medicine",
"and",
"health",
"sciences",
"infectious",
"disease",
"epidemiology",
"epidemiology",
"plant",
"pathology",
"biology",
"and",
"life",
"sciences",
"population",
"biology",
"evolutionary",
"biology",
"evolutionary",
"systemat... | 2014 | Inference of Epidemiological Dynamics Based on Simulated Phylogenies Using Birth-Death and Coalescent Models |
Asexual spores ( conidia ) are the infectious propagules of many pathogenic fungi , and the capacity to sense the host environment and trigger conidial germination is a key pathogenicity determinant . Germination of conidia requires the de novo establishment of a polarised growth axis and consequent germ tube extension . The molecular mechanisms that control polarisation during germination are poorly understood . In the dimorphic human pathogenic fungus Penicillium marneffei , conidia germinate to produce one of two cell types that have very different fates in response to an environmental cue . At 25 °C , conidia germinate to produce the saprophytic cell type , septate , multinucleate hyphae that have the capacity to undergo asexual development . At 37 °C , conidia germinate to produce the pathogenic cell type , arthroconidiating hyphae that liberate uninucleate yeast cells . This study shows that the p21-activated kinase pakA is an essential component of the polarity establishment machinery during conidial germination and polarised growth of yeast cells at 37 °C but is not required for germination or polarised growth at 25 °C . Analysis shows that the heterotrimeric G protein α subunit GasC and the CDC42 orthologue CflA lie upstream of PakA for germination at both temperatures , while the Ras orthologue RasA only functions at 25 °C . These findings suggest that although some proteins that regulate the establishment of polarised growth in budding yeast are conserved in filamentous fungi , the circuitry and downstream effectors are differentially regulated to give rise to distinct cell types .
The generation of an axis of cell polarity is central to the activity of many cells and the establishment of a wide variety of cell morphologies . It relies on the ability to mark different cellular regions by specific protein localisation . The establishment of polarised growth requires selection of a site to which proteins and components of the cytoskeleton are recruited . Growth is then directed specifically to this site via targeted cellular trafficking and concomitant cell wall deposition . Fungi are small eukaryotes that exhibit highly polarised growth patterns and provide excellent models for the study of the molecular mechanisms underlying cell polarity . Saccharomyces cerevisiae establishes a polarised axis of growth during the processes of budding cell division , schmoo formation during mating , and pseudohyphal growth . Under conditions of nitrogen starvation , S . cerevisiae diploid cells undergo pseudohyphal growth , a morphological switch that involves changes in cell shape and division [1] . Pseudohyphal growth requires the initiation of polarised growth for cellular elongation and is under the control of two signaling pathways: a cyclic adenosine monophosphate ( cAMP ) /protein kinase A ( PKA ) pathway and a mitogen-activated protein kinase ( MAPK ) cascade . The cAMP/PKA pathway is activated by a glucose/sucrose sensitive receptor Gpr1p , which activates the G protein Gpa2p ( α subunit ) , which in turn is inhibited by the novel kelch-Gβ subunits Gpb1/2p , a third subunit Gpg1p , and a negative regulator Rgs2p [2–8] . The Gpa2-Gpb1/2 complex regulates the cAMP/PKA pathway directly in association with the Ras2p GTPase and the RasGAP neurofibromin homologues Ira1/2p [6–7 , 9] . Ras2p can also activate the MAPK cascade by activating the guanine nucleotide exchange factor Cdc24p , which catalyzes the guanosine diphosphate ( GDP ) to guanosine triphosphate ( GTP ) exchange of the Rho GTPase Cdc42p [1 , 10 , 11] . GTP-bound Cdc42p is required to initiate actin polarisation and recruits and activates additional proteins required for polarised growth such as septins , myosins , and the p21-activated kinase ( PAK ) Ste20p ( reviewed in [12] ) . In turn , Ste20p activates the MAPK cascade by phosphorylating Ste11p ( MAPKKK ) ; Ste11p phosphorylates Ste7p ( MAPKK ) , which then activates the Kss1p MAPK [13–15] . Filamentous fungi exhibit a highly polarised axis during vegetative hyphal growth , asexual ( conidiation ) and sexual ( mating ) development , and initiation of polarised filamentous growth during the germination of asexual ( conidia ) and sexual ( ascospores ) spores . Spherical conidia germinate under favorable environmental conditions by initially growing isotropically . This growth is followed by the establishment of a de novo axis of polarised growth to allow a germ tube to emerge [16–18] . Conidial germination is a central aspect of fungal cell propagation , initiating the formation of the extensive radiating hyphal network necessary for colonization of substrates from a dormant conidium . Conidia are also the infectious propagules of many pathogenic fungi , and germination of conidia in the lung or leaves of potential hosts is likely to be a key pathogenicity determinant . Studies in the filamentous fungus Aspergillus nidulans have shown that the cAMP-PKA and Ras pathways play a role [18–20] . However , the molecular mechanisms governing germination of fungal conidia are not well understood in any system . Penicillium marneffei is an opportunistic human pathogen with a thermally regulated dimorphic switch . At 25 °C , in the saprophytic growth phase , conidia germinate to produce highly polarised , septate , branched , multinucleate hyphae . Conidia can also germinate at 37 °C to produce polarised arthroconidiating hyphae , in which nuclear division and septation are coupled , double septa are laid down , and fragmentation occurs along this plane to liberate uninucleate yeast cells that consequently divide by fission [21] . The yeast cells are the pathogenic growth form and multiple yeast cells are observed in the pulmonary alveolar macrophages of infected individuals . P . marneffei infection is thought to occur through inhalation of conidia , which bind to the laminin in the broncholalvelolar epithelia . Conidia are then ingested by pulmonary alveolar macrophages and germinate , generating the uninucleate yeast cells [21] . Therefore in P . marneffei , conidial germination can lead to two very different morphological programs in response to different temperatures . Some of the core components regulating polarised growth establishment in S . cerevisiae are conserved during polarity establishment in germinating conidia of filamentous fungi . The P . marneffei Gpa2p homologue , encoded by gasC , is required during conidial germination to produce hyphae . Deletion of gasC results in delayed germination , whereas expression of a dominant activating allele shows a significantly accelerated germination rate [22] . Likewise , the P . marneffei Ras homologue , RasA , is required for conidial germination where expression of either a dominant negative or activated allele results in a germination delay and conidia with abnormal isotropic growth [23] . Expression of either a dominant negative or constitutively active allele of the P . marneffei CDC42 homologue ( cflA ) results in a decrease or increase in the rate of germination at 25 °C , respectively [24] . The role of GasC , RasA , and CflA during conidial germination in P . marneffei suggests that the core components regulating polarised growth establishment in S . cerevisiae may be conserved during polarity establishment in germinating conidia of filamentous fungi . To investigate if any of the downstream effectors of this core pathway are also conserved in function , a PAK STE20 homologue ( designated pakA ) was identified and deleted in P . marneffei . Characterization of pakA in P . marneffei has shown that this gene is essential for conidial germination at 37 °C and for polarised growth of yeast cells and is downstream of both a heterotrimeric G protein and Cdc42 pathway . In contrast , PakA plays only a minor role during germination of conidia at 25 °C and is not required for polarised growth of hyphae . Germination in this case is controlled by a heterotrimeric G protein–Ras-Cdc42 pathway . These data suggest that although some proteins that regulate the establishment of polarised growth in budding yeast and filamentous fungi may be conserved , the downstream effectors are likely to be different or regulated differently to give rise to the distinct cell types in these two modes of growth .
An A . nidulans sequence was identified from the genome sequence ( http://www . broad . mit . edu/annotation/genome/aspergillus_group/MultiHome . html ) with strong homology to Candida albicans Cst20p ( 67% identity , 81% similarity ) and other PAKs , including S . cerevisiae Ste20p ( 71% identity , 81% similarity , accession number AAA35039 ) . Primers were designed to amplify the sequence encoding the conserved kinase domain and a PCR product was generated using A . nidulans genomic DNA . The A . nidulans STE20 homologous sequence was used to screen a P . marneffei genomic library at low stringency . Five positive clones were identified , which fell into two classes based on restriction enzyme digestion patterns . A 6 . 4 kb NotI/BglII hybridizing fragment from one of these classes was subcloned into NotI/BamHI digested pBluescript II SK+ ( pKB5751 ) . Sequencing revealed strong sequence homology to STE20-like PAKs from Magnaporthe grisea ( 78% identity , 89% similarity , accession number AAP93639 ) , Ustilago maydis ( 74% identity , 83% similarity , accession number AAM97788 ) , and S . cerevisiae ( 72% identity , 83% similarity ) . The gene within this clone was subsequently named pakA . The pakA open reading frame spans 2407 bp and contains seven exons and six introns . The predicted protein is 642 amino acids in length and contains a Cdc42/Rac interactive binding ( CRIB ) domain at positions 98–115 , a predicted kinase domain at 361–612 , and a Gβ binding domain at 619–629 . Preliminary analysis of the second class of clones revealed that the gene within these clones has strong sequence homology to CLA4-like PAKs . RNA was isolated from vegetative hyphae grown for 2 d in liquid medium at 25 °C , asexual development ( conidiation ) cultures grown for 4 d on solid medium at 25 °C , and yeast cells grown for 8 d in liquid medium at 37 °C . A pakA transcript was detected under all conditions . The amount of pakA transcript was approximately equivalent under all conditions when compared with the histone H3 control ( unpublished data ) . A pakA construct , which deleted from −425 to +2030 of pakA , was linearised and transformed into P . marneffei strain SPM4 ( niaD1 pyrG1 ) and pyrG+ transformants selected . PyrG+ transformants were screened by genomic Southern blotting and one strain was isolated that possessed a restriction pattern consistent with replacement of pakA by pyrG at the genomic locus . The deletion strain was plated on medium containing 5-fluoroorotic acid ( 5-FOA ) to generate a ΔpakA pyrG strain ( Materials and Methods ) . This strain was cotransformed with plasmids containing pakA+ and pyrG+ genes and co-transformants confirmed by Southern blot analysis . The transformants contained from 1–7 copies of pakA . In S . cerevisiae , the H345G mutation in the conserved CRIB domain of Ste20p results in a mutant protein that shows reduced interaction with the upstream activator Cdc42p ( GTPase ) in a two-hybrid assay and loss of correct localisation to sites of polarised growth [25] . The phenotype of the STE20H345G strain is almost equivalent to the null , indicating that the interaction of Cdc42p and Ste20p is essential for Ste20p function [25] . The equivalent mutation was generated in P . marneffei pakA ( pakAH108G ) by inverse PCR ( Materials and Methods ) . The pakAH108G construct was co-transformed with pyrG+ into the ΔpakA pyrG− strain and transformants selected for PyrG+ . Co-transformation was confirmed by Southern blot analysis of genomic DNA and four representative transformants with copy numbers ranging from 3 to 12 were selected for further analysis . The gfp::pakA and gfp::pakAH108G fusion constructs were generated and co-transformed with the pyrG+ gene into the ΔpakA pyrG strain to investigate the localisation of PakA and to assess whether the pakAH108G mutation affects PakA localisation . Transformants were selected for PyrG+ and confirmed by Southern blot analysis of genomic DNA . Four transformants of each genotype were selected for further analysis and had copy numbers ranging from 4 to 9 for gfp::pakA and from 2 to 20 for gfp::pakAH108G . The gfp::pakA and gfp::pakAH108G strains were grown on agar-coated slides for 2 and 4 d at 25 °C and 37 °C , respectively . At 25 °C , the GFP::PakA fusion protein was concentrated at the hyphal apex ( Figure 1A ) and localised as spots along the subapical cells of the hyphae ( Figure 1B ) . In contrast , the GFP::PakAH108G fusion protein was visible as diffuse fluorescence in the cytoplasm but not concentrated at the hyphal apex or as spots along the hyphae ( Figure 1A and 1B ) . At 37 °C , the GFP::PakA fusion protein was concentrated at the apex of arthroconidiating hyphae , although this fluorescence was less intense than that visualized in the same transformants at 25 °C . The GFP::PakAH108G fusion protein was not observed at the apex of arthroconidiating hyphae ( unpublished data ) . To investigate whether the GFP::PakA fusion protein co-localises with actin at the hyphal apex and at nascent septation sites , immunostaining using mouse anti-actin and rabbit anti-GFP antibodies was performed on two of the gfp::pakA and two of the gfp::pakAH108G strains at both 25 °C and 37 °C . At 25 °C , actin is localised as cortical actin spots along the hyphae and is concentrated at nascent septation sites and the hyphal apex ( Figure 1C–1F ) . The GFP::PakA fusion protein co-localised with actin at all of these locations ( Figure 1C–1F ) . The GFP::PakAH108G fusion protein also co-localised with actin at cortical actin patches , nascent septation sites , and the hyphal apex , in addition to showing diffuse staining throughout the cytoplasm ( Figure 1C–1F ) . At 37 °C , actin is also localised as cortical patches along the arthroconidiating hyphae , concentrated at nascent septation sites and the apex of arthroconidiating hyphae ( Figure S1 ) . The GFP::PakA and GFP::PakAH108G fusion proteins co-localised with actin at all of these sites ( Figure S1 ) . At 25 °C , P . marneffei colonies are comprised of highly polarised vegetative hyphae growing on the agar surface and bearing asexual structures ( conidiophores ) that appear green due to the presence of pigmented asexual spores ( conidia ) on the conidiophores . After 5 d growth at 25 °C , surface hyphae are visible in the wild-type strain ( Figure 2A ) . The ΔpakA pakA+ strains appeared wild-type after 5 d at 25° C , whereas both the ΔpakA and ΔpakA pakAH108G strains showed a reduction in growth ( Figure 2A ) . Despite the initial reduction in growth , after 10 d all strains were producing conidia . P . marneffei colonies are yeast-like at 37 °C and ΔpakA pakA+ strains were indistinguishable from the wild type when grown for 4 d at 37 °C ( Figure 2B ) . In contrast , the ΔpakA and ΔpakA pakAH108G strains displayed reduced growth at 37 °C ( Figure 2B ) . Both the ΔpakA and ΔpakA pakAH108G strains exhibit a delay in growth after 5 d at 25 °C ( Figure 2A ) . One explanation for this difference could be a delay in the germination of conidia . The kinetics of germination were measured at 25 °C in the wild-type ( pakA+ ) , ΔpakA , ΔpakA pakA+ , and two ΔpakA pakAH108G strains by counting the number of ungerminated versus germinated conidia ( conidia with a visible germ tube ) in a population of 100 in three independent experiments after 8 , 12 , or 15 h incubation in liquid media ( Table S1 ) . The complemented ΔpakA strain ( ΔpakA pakA+ ) is indistinguishable from wild type ( Figure 3A and 3B ) . The ΔpakA and ΔpakA pakAH108G strains show a minor delay in germination at 25 °C ( Figure 3A and 3B ) . To investigate if the deletion of pakA or the pakAH108G mutation results in aberrant hyphal morphology or asexual development at 25 °C the wild-type , ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains were grown on agar-coated slides for 2 or 4 d at 25 °C , stained with calcofluor ( to observe cell walls ) or 4′6-diamidino-2-phenylindole ( DAPI; to observe nuclei ) , and examined microscopically . After 2 d at 25 °C , wild-type P . marneffei grows as septate , branched hyphae of which subapical cells are predominately uninucleate and apical cells are multinucleate . The ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains were indistinguishable from wild type after 2 d , with normal morphology , septation , branching , and nuclear index . After 4 d at 25 °C , wild-type P . marneffei begins to undergo asexual development , with the production of a specialized stalk from which differentiated cells are produced sequentially in a budding fashion: metulae bud from the stalk , phialides bud from metulae , and uninucleate conidia bud from phialides . The ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains produced conidiophores with wild-type morphology ( unpublished data ) . In contrast to the wild-type and ΔpakA pakA+ strains , the ΔpakA and ΔpakA pakAH108G strains displayed reduced growth rates at 37 °C ( Figure 2B ) . To assess if the basis of this difference is because pakA is required during the germination of conidia or during yeast morphogenesis and growth at 37 °C , the wild-type ( pakA+ ) , ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains were inoculated on agar-coated slides and incubated for 4 d at 37 °C . It was immediately apparent that the ΔpakA and ΔpakA pakAH108G strains possessed a severe germination defect and almost all of the conidia remained ungerminated . The germination kinetics were measured by counting the number of ungerminated versus germinated conidia ( conidia with a visible germ tube ) in a population of 100 in three independent experiments after 8 , 12 , 15 , or 20 h in liquid medium ( Table S2 ) . Germination is slower and germlings appear fatter at 37 °C compared with 25 °C ( Figure 3A and 3C ) . In contrast to wild-type and the ΔpakA pakA+ strains , both the ΔpakA strain and the ΔpakA pakAH108G strains showed a severe defect in germination ( Figure 3C and 3D ) . Despite the majority of conidia remaining ungerminated in the ΔpakA and ΔpakA pakAH108G strains , a small proportion do germinate , and it is presumably these cells that go on to establish the colony . Growth of the pakA strains in liquid medium showed that the small proportion of conidia that germinate go on to form arthroconidiating hyphae that fragment at septation sites to liberate uninucleate yeast cells , just like the wild type . The wild-type ( pakA+ ) , ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains were grown in liquid brain heart infusion ( BHI ) for 6 d at 37 °C and cells were stained with calcofluor or DAPI and observed microscopically ( Figure 4 ) . In wild type , the culture consists of a mixture of fragmented arthroconidiating hyphae and uninucleate yeast cells . The ΔpakA strain produced swollen arthroconidiating hyphae and yeast cells with increased septation ( Figure 4 ) but normal nuclear index . These defects were complemented when the strain was transformed with pakA+ . The ΔpakA pakAH108G strains also produced swollen arthroconidiating hyphae and yeast cells , but the phenotype was more severe than the ΔpakA strain , with very few yeast cells produced and—in contrast to the ΔpakA strain and wild type—a decrease in the septation index ( Figure 4 ) . Therefore , the pakAH108G allele has an inhibitory activity on septation . P . marneffei infection occurs through the inhalation of conidia . The conidia are ingested by host pulmonary alveolar macrophages where they germinate into unicellular yeast cells that divide by fission . Multiple yeast cells are seen in the pulmonary alveolar macrophages and peripheral blood mononuclear cells of infected individuals [21] . To investigate if perturbing conidial germination will influence pathogenicity , the ability of the mutant pakA strains to germinate into pathogenic yeast cells was investigated . LPS activated J774 murine macrophages were infected with no conidia or with conidia of the wild-type ( pakA+ ) , ΔpakA , ΔpakA pakAH108G , or ΔpakA pakA+ strains ( Materials and Methods ) . After 24 h , numerous yeast cells dividing by fission were observed in macrophages infected with wild-type ( pakA+ ) or ΔpakA pakA+ conidia ( Figure 5 ) . Only 16 . 3 ± 2 . 50% or 21 . 5 ± 2 . 07% of wild-type ( pakA+ ) or ΔpakA pakA+ conidia , respectively , remained ungerminated in infected macrophages . In contrast , conidia of the ΔpakA or ΔpakA pakAH108G strains remained predominately ungerminated in infected macrophages ( Figure 5 ) . 60 . 0 ± 3 . 80% of ΔpakA conidia and 66 . 1 ± 7 . 16% of ΔpakA pakAH108G conidia remained ungerminated in macrophages 24 h post infection . This suggests that PakA is required for conidial germination during infection of host macrophages . The temperature-dependent regulation of conidial germination could be the result of the presence of a temperature-specific factor on which PakA depends or the result of a change in the thermostability of a complex in which PakA operates . To distinguish between these two possibilities , wild-type and ΔpakA conidia were incubated in liquid medium at different temperatures ranging from 25 °C to 37 °C , in 2 °C increments , and germination rates were measured ( Materials and Methods ) . In contrast to the wild type , the ΔpakA strain showed a gradual decrease in the percentage of germination as the temperature increased , indicating that there is no critical temperature during the switch ( Figure 3E ) . This result supports the latter hypothesis and suggests PakA-dependent thermosensitivity . To investigate if ΔpakA conidia at 37 °C are waiting for the signal to germinate or have aborted , ΔpakA conidia were incubated at 37 °C for 20 h , then at 25 °C for 20 h . The majority of conidia germinated upon switching to 25 °C , indicating that after 20 h incubation at 37 °C , the ΔpakA conidia remained viable ( unpublished data ) . To identify other factors involved in the pakA-dependent differences in germination at 25 °C and 37 °C , conidial germination was analyzed in detail at both temperatures in strains carrying mutations in cflA ( CDC42 orthologue ) , gasC ( GPA2 orthologue ) , and rasA ( RAS2 orthologue ) , which have been shown previously to affect germination at 25 °C but which had not been characterized at 37 °C [22–24] . The role of CflA , GasC , and RasA in germination at 37 °C was characterized by assessing the percentage of germinated conidia after 8 , 12 , 15 , and 20 h at both 25 °C and 37 °C in two strains of each genotype ( Tables S1 and S2 ) . Two-level nested ANOVA was performed on the data for each time point at both 25 °C and 37 °C to test if germination rates differed significantly between genotypes and also between transformants of the same genotype ( Materials and Methods ) . ANOVA showed there was a significant difference between genotypes in all time points except 0 h . In a few instances , there was variation within transformants of the same genotype ( Tables S1 and S2 ) . Strains expressing the dominant negative cflAD120A allele showed delayed germination at both 25 °C and 37 °C ( Figure 6A and 6C ) . Dominant activated cflAG14V strains displayed accelerated germination at both 25 °C and 37 °C ( Figure 6B and 6D ) . These results indicate that active CflA promotes conidial germination at both 25 °C and 37 °C . Likewise , the dominant interfering gasCG207R strains showed delayed germination at both 25 °C and 37 °C ( Figure 7A and 7C ) . The dominant activated gasCG45R strains showed accelerated germination at 25 °C and 37 °C , suggesting that , like CflA , GasC is required for conidial germination at both 25 °C and 37 °C ( Figure 7B and 7D ) . Both the dominant activated ( rasAG19V ) and dominant negative ( rasAD125A ) rasA strains showed a decrease in germination at 25 °C ( Table S1 ) . In contrast , both the dominant negative and dominant activated strains showed wild type germination patterns at 37 °C , suggesting that RasA is required for conidial germination at 25 °C but not at 37 °C ( Table S2 ) . To investigate any genetic interaction between pakA and cflA , double mutants were generated ( ΔpakA cflA+ , ΔpakA cflAD120A , ΔpakA cflAG14V , ΔpakA pakAH108G cflA+ , ΔpakA pakAH108G cflAD120A , and ΔpakA pakAH108G cflAG14V ) and germination was characterized by assessing the percentage of germinated conidia after 8 , 12 , 15 , and 20 h at both 25 °C and 37 °C ( Tables S1 and S2 ) . It should be noted that multiple copy integrants may result in significant overexpression . Two strains of each genotype were assessed and a single representative strain is shown in Figure 6 . ANOVA was performed on the data for each time point at both 25 °C and 37 °C to test if germination rates differed significantly between genotypes and also between transformants of the same genotype ( Materials and Methods ) . ANOVA showed there was a significant difference between genotypes at all time points except 0 h . In a few instances , there was variation within transformants of the same genotype ( Tables S1 and S2 ) . The control ΔpakA cflA+ and ΔpakA pakAH108G cflA+ strains showed germination patterns at 25 °C and 37 °C that were indistinguishable from the parental ΔpakA and ΔpakA pakAH108G strains . At 25 °C , the ΔpakA cflAD120A and ΔpakA pakAH108G cflAD120A strains displayed delayed conidial germination at rates similar to the single cflAD120A mutant strains ( Figure 6A ) . At 37 °C , the ΔpakA cflAD120A and ΔpakA pakAH108G cflAD120A strains displayed dramatically reduced germination like the single ΔpakA and ΔpakA pakAH108G strains ( Figure 6C ) . In contrast to the accelerated germination observed in cflAG14V mutants at 25 °C , which is much faster than wild-type ( Figure 6B ) , the ΔpakA cflAG14V and ΔpakA pakAH108G cflAG14V double mutants display slower than wild-type germination at 25 °C , similar to the ΔpakA and ΔpakA pakAH108G single mutant strains ( Figure 6B ) . This suggests that the accelerated germination observed in cflAG14V strains at 25 °C requires active PakA and an interaction of CflA and PakA via the PakA CRIB domain . Likewise at 37 °C , in contrast to the accelerated germination observed in cflAG14V mutants , the ΔpakA cflAG14V and ΔpakA pakAH108G cflAG14V double mutants display dramatically reduced germination at rates equivalent to the ΔpakA and ΔpakA pakAH108G single mutant strains ( Figure 6D ) . The inability of the cflAG14V dominant activated allele to suppress the reduced germination phenotype of the ΔpakA and ΔpakA pakAH108G mutants suggests that at 37 °C , CflA acts upstream of PakA during germination and that an interaction between CflA and the CRIB domain of PakA is required for germination to proceed . P . marneffei RasA operates upstream of CflA at both 25 °C and 37 °C [23] . The genetic interaction of pakA and rasA was investigated by generating ΔpakA rasA+ , ΔpakA rasAD125A , and ΔpakA rasAG19V double mutants . At 25 °C , the ΔpakA rasA+ strains showed wild-type germination patterns , whereas the ΔpakA rasAD125A and ΔpakA rasAG19V mutants showed delayed germination similar to the single rasAD125A and rasAG19V mutants ( Table S1 ) . At 37 °C , the ΔpakA rasA+ , ΔpakA rasAD125A , and ΔpakA rasAG19V double mutants showed severely reduced germination similar to the ΔpakA mutant ( Table S2 ) . To investigate the genetic interaction between pakA and gasC , double mutants were generated ( ΔpakA gasC+ , ΔpakA gasCG207R , and ΔpakA gasCG45R ) . Germination was characterized in five strains of each double mutant genotype by assessing the percentage of germinated conidia after 8 , 12 , 15 , and 20 h at both 25 °C and 37 °C ( Tables S1 and S2 ) . ANOVA was performed on the data for each time point at both 25 °C and 37 °C ( Materials and Methods ) and revealed there was a significant difference between genotypes at all time points except 0 h . In a few instances , there was variation within transformants of the same genotype ( Tables S1 and S2 ) . ΔpakA gasC+ strains were indistinguishable from ΔpakA . At 25 °C , ΔpakA gasCG207R strains have delayed germination , which is slower than both the gasCG207R single mutant strains and the ΔpakA mutant ( Figure 7A ) . At 37 °C ΔpakA gasCG207R strains have a severe reduction in germination similar to ΔpakA ( Figure 7C ) . In contrast to the accelerated germination of the gasCG45R single mutant strains and the delayed germination of the ΔpakA mutant , the ΔpakA gasCG45R double mutant strains show wild-type germination at 25 °C ( Figure 7B ) . In addition , ΔpakA gasCG45R double mutant strains show germination rates at 37 °C , which are lower than those of wild-type and the gasCG45R mutants but higher than that of the ΔpakA mutant ( Figure 7D ) . This indicates that expression of the gasCG45R mutant allele partially suppresses the germination defects of ΔpakA and suggests that GasC regulates two pathways during germination , one of which is independent of PakA .
The establishment of an axis of polarised growth is orchestrated by the asymmetric distribution of cellular components through the localisation and activation of proteins required for growth . Some of the core components regulating polarised growth establishment in S . cerevisiae are conserved both in the genome and functionally during polarity establishment in more complex organisms . However , the question remains as to how multi-cellular organisms generate the greater diversity of distinct cell types with the same set of core components . Unlike small eukaryotes like fungi , larger eukaryotes such as flies and mammals often have an increased number of factors involved in polarity establishment with significant redundancy [26–28] . One possible mechanism is to alter the activity of the key establishment proteins in different cell types , while another is to differentially regulate the effector proteins . In the dimorphic pathogen P . marneffei , the germination of conidia gives rise to two different developmental pathways and cell types . The regulation of conidial germination by CflA ( CDC42 orthologue ) and GasC ( GPA2 orthologue ) at both 25 °C and 37 °C suggests that the core components regulating polarised growth establishment in S . cerevisiae may be conserved during polarity establishment in germinating conidia of filamentous fungi . However , mutations in pakA , a potential downstream effector of cflA , result in a dramatic reduction in the rates of germination at 37 °C but not 25 °C . This suggests that in P . marneffei conserved polarity establishment proteins regulate germination , but the downstream effectors are differentially regulated to give rise to distinct cell types . The results suggest a model in which , at 25 °C , GasC activates two pathways regulating conidial germination ( Figure 8 ) . In one pathway , RasA activates CflA , which activates PakA—by association via the CRIB domain—and a proposed additional effector to establish polarised hyphal growth ( Figure 8 ) . At 37 °C , GasC also activates two pathways regulating conidial germination . In one pathway , similar to 25 °C , CflA activates PakA via the CRIB domain , and this interaction is required for PakA function . Unlike 25 °C , RasA does not activate CflA , PakA plays a crucial role during the establishment of polarised arthroconidiating hyphal growth , and no additional effector is required ( Figure 8 ) . gasC encodes a heterotrimeric guanine nucleotide-binding ( G-protein ) α-subunit with homology to S . cerevisiae GPA2 . Gpa2p is required for the initiation of filamentous growth ( pseudohyphal growth ) in response to nitrogen starvation via the activation of the cAMP-PKA pathway [2 , 3] . In P . marneffei , expression of dominant activated gasCG45R and dominant interfering gasCG207R alleles at both 25 °C and 37 °C results in accelerated or delayed conidial germination , respectively , and suggests that GasC acts as a general upstream activator initiating filamentous growth . This study suggests that GasC activates two pathways regulating germination at both 25 °C and 37 °C , one of which is dependent on PakA ( Figure 8 ) . In contrast to the accelerated germination of gasCG45R strains and the slightly delayed germination of ΔpakA , ΔpakA gasCG45R double mutant strains display wild-type germination at 25 °C . A reduction in the germination rates suggests that PakA is acting downstream of GasC at 25 °C . However , as germination of the ΔpakA gasCG45R strains is higher than ΔpakA , GasC must also regulate an additional PakA-independent pathway activating germination at 25 °C . This hypothesis is supported by the germination rates observed in the ΔpakA gasCG207R strains , which are lower than both the single mutant strains , suggesting these mutations have an additive effect . Likewise , at 37 °C , partial suppression of the ΔpakA germination defect by expression of the gasCG45R allele indicates that GasC acts as an upstream regulator of two pathways activating conidial germination , one of which is dependent on PakA . The activation of two pathways regulating the initiation of filamentous growth in P . marneffei differs from S . cerevisiae . In S . cerevisiae , the Gpa2p regulation of pseudohyphal growth occurs via the activation of the cAMP-PKA pathway , which is independent of the Ste20p/MAPK cascade [3] . The pseudohyphal growth defect of a gpa2 mutant can be suppressed by addition of cAMP or overexpression of the dominant activated RAS2G19V allele . However , the pseudohyphal defect cannot be suppressed by the overexpression of the dominant active STE11–4 allele and has no effect on expression of a reporter gene ( FRE-lacZ ) , which is known to be regulated by the MAP kinase cascade [3] . These results together suggest that , unlike P . marneffei GasC , S . cerevisiae , Gpa2p regulates the cAMP pathway but not the Ste20p-MAPK cascade , thus regulating the initiation of filamentous growth . In S . cerevisiae Cdc42p localises and activates Ste20p ( reviewed in [12] ) . Therefore P . marneffei PakA is a potential downstream effector of the CflA ( Cdc42p orthologue ) and may be involved in similar processes . Like cflAD120A strains , which show delayed germination at 37 °C , the ΔpakA mutant exhibits a dramatic reduction in conidial germination at 37 °C , suggesting that PakA is crucial during the establishment of polarised growth to give rise to arthroconidiating hyphae . However , unlike CflA , PakA plays only a minor role during germination at 25 °C , as strains expressing the dominant negative cflAD120A allele show a severe delay in germination compared with only a slight delay in conidial germination for the ΔpakA and ΔpakA pakAH108G strains . These results suggest that PakA acts downstream of CflA at both 25 °C and 37 °C in P . marneffei . This hypothesis is also supported by the observation that the accelerated germination seen in cflAG14V strains is abrogated by the ΔpakA at both 25 °C and 37 °C in ΔpakA cflAG14V strains . PAKs contain a conserved N-terminal CRIB domain and a C-terminal kinase domain ( reviewed in [27] ) . The CRIB domain of S . cerevisiae Ste20p negatively inhibits the kinase domain preventing signaling [25] . This autoinhibition is relieved by interaction of the CRIB domain with Cdc42p , and this interaction is also necessary for localisation of Ste20p to sites of polarised growth [25] . The H345G mutation in the Ste20p CRIB domain shows reduced interaction with Cdc42p in a two-hybrid assay and the loss of correct localisation to sites of polarised growth . The phenotype of the STE20H345G strain is almost equivalent to the null , indicating that the interaction of Cdc42p and Ste20p is essential for Ste20p function ( and for relief of autoinhibition of the kinase domain ) [25] . The H108G CRIB domain mutation in P . marneffei is equivalent to the H345G of S . cerevisiae and was found to have a similar effect . The H108G mutation resulted in reduced localisation of PakA to sites of polarised growth and a phenotype equivalent to the deletion mutant . Compared with the accelerated germination of cflAG14V single mutant strains at both 25 °C and 37 °C , the ΔpakA pakAH108G cflAG14V mutant strains exhibited reduced germination , suggesting that the interaction of CflA and PakA via the CRIB domain is required during conidial germination at both 25 °C and 37 °C . The interaction of CflA and PakA is also required during polarised growth of yeast cells , as the ΔpakA pakAH108G strains showed a swollen , abnormal yeast morphology at 37 °C , similar to the deletion strain . In S . cerevisiae , the CRIB domain is essential for pseudohyphal growth but dispensable for G protein–mediated pheromone signaling [11] . In P . marneffei the CRIB domain of PakA is also required for the initiation of filamentous growth , but unlike S . cerevisiae the initiation of filamentous growth in P . marneffei by PakA is partially dependent on G-protein signaling . The minor role played by PakA during germination and hyphal growth at 25 °C suggests that another CflA effector , possibly a Cla4p orthologue , is required for these processes . The genomes of A . nidulans , M . grisea , U . maydis , Coprinopsis cinerea , Neurospora crassa , and C . albicans encode two PAKs , one with homology to Ste20p and the other to Cla4p ( http://www . broad . mit . edu/annotation/fgi/ ) . In addition to the CRIB and kinase domains of Ste20p orthologues , Cla4p homologues have a pleckstrin homology domain . Cla4p in S . cerevisiae is required for septation but does not play a role in pseudohyphal growth . However , deletion of CLA4 in the yeasts Yarrowia lipolytica and C . albicans blocks filament formation , and the cla4 deletion mutant of the plant pathogen U . maydis is unable to form filaments during infection [29–32] . cflA has been previously shown to play a pivotal role during hyphal morphogenesis with mutations resulting in grossly aberrant hyphae [24] . It was therefore expected that the ΔpakA strain may have a similar phenotype , albeit less severe , as CflA is proposed to interact and activate numerous effector proteins . The lack of a ΔpakA hyphal phenotype suggests that PakA is not required for hyphal growth . Like P . marneffei , deletion of the Ste20p homologues in C . albicans and U . maydis does not result in defects in hyphal morphology [33 , 34] . However , the co-localisation of the GFP::PakA fusion protein with actin and the localisation to the same cellular locations as CflA at nascent septation sites and to the hyphal apex suggests a role during polarised growth of hyphae [23] . In S . cerevisiae , Ste20p directly phosphorylates Bni1p , a component of the polarisome [35] . The polarisome is a protein complex , which contains Bni1p , Spa2p , Pea2p , and Bud6p , that promotes polarised morphogenesis during filamentous growth [35 , 36] . The Bni1p homologue , SepA , plays a conserved role in polarised growth in A . nidulans [37] . However , A . nidulans lacks a Pea2p homologue and the Spa2 homologue , SpaA , is only partially conserved in sequence and function , indicating that the polarisome in filamentous fungi likely consists of a modified set of components with different contributions to polarisome function [36] . The implication is that in P . marneffei , in addition to specific developmental roles , pakA and pakB play complementary , and possibly overlapping , roles in the establishment of polarised growth during conidial germination and in the maintenance of an axis of polarisation during hyphal growth . How the two PAKs coordinately regulate different aspects of development of multi-cellular fungi still remains unclear . The analysis of a CLA4 orthologue from P . marneffei may resolve many of these issues .
P . marneffei genomic DNA and RNA was isolated as previously described [38 , 39] . Southern and northern blotting was performed with Amersham Hybond N+ membrane according to the manufacturer's instructions . Filters were hybridized using [α−32P]dATP-labeled probes by standard methods [40] . Primers L18 ( 5′-TGATCCCACAAAACTTTACT-3′ ) and L19 ( 5′-GCTCGTTTCTCAGGGTCCAC-3′ ) were used to amplify the A . nidulans genomic sequence encoding the conserved kinase domain of the STE20 homologue . The PCR product was sequenced and used to screen a P . marneffei genomic library ( constructed in λGEM-11 ) at low stringency ( 50% formamide , 2 x SSC , 37 °C ) . A 6 . 4 kb NotI/BglII hybridizing fragment from a positively hybridizing clone was subcloned into NotI/BamHI digested pBluescript II SK+ ( pKB5751 ) . Sequencing was performed by the Australian Genome Research Facility and analyzed using Sequencher 3 . 1 . 1 ( Gene Codes Corporation ) . The Genbank accession number of the P . marneffei pakA gene is AY621630 . A pakA deletion construct ( pKB5792 ) was generated by replacing the 2 . 5 kb EcoRV/ClaI fragment of pKB5751 with the 2 . 5 kb SmaI/ClaI fragment containing the pyrG+ selectable marker . This resulted in pyrG+ flanked by 2 . 6 kb of 5′ and 1 . 1 kb of 3′ pakA sequence , and deleted from −425 to +2030 . Inverse PCR using the mutagenic primers N30 ( 5′-ACATGAGTAACACCGACAGGG-3′ ) and N32 ( 5′-TGGATACGACAATCAGACTGG-3′ ) was used to introduce the H108G mutation into pakA generating pKB5908 . The integrity of the construct was confirmed by sequencing . The gfp::pakA and gfp::pakAH108G constructs were generated by ligating a BamHI/XbaI fragment from pKB5751 ( pakA ) and pKB5908 ( pakAH108G ) into pALX196 ( gpdA ( p ) ::gfp ) . Strains used in this study are shown in Table 1 . The ΔpakA strain ( ΔpakA::pyrG+ ) was generated by transforming the strain SPM4 with linearised pKB5792 and selecting for pyrG+ . Transformation was performed using the previously described protoplast method [38] . The ΔpakA pyrG− strain was isolated by plating the ΔpakA strain ( ΔpakA::pyrG+ ) on medium containing 1 mg/mL−1 5-FOA supplemented with 10 mM γ-amino butyric acid ( GABA ) and 5 mM uracil to select for the loss of the pyrG marker . A 5-FOA resistant sector was isolated that had a restriction pattern consistent with loss of pyrG at the pakA locus . The strain is unable to grow in the absence of 5 mM uracil . P . marneffei FRR2161 , SPM4 , cflAD120A , cflAG14V , rasAD125A , rasAG19V , gasCG207R , and gasCG45R have been previously described [22–24 , 38] . All other strains listed in Table 1 were generated by cotransformation of the ΔpakA pyrG or ΔpakA strain with plasmids containing the appropriate mutant allele and either pAB4342 ( pyrG+ ) or pMT1612 ( barA+ ) as selectable markers . Southern blot analysis was used to confirm cotransformation and to determine the plasmid copy number . At 25 °C strains were grown on A . nidulans minimal medium ( ANM ) supplemented with 1% glucose and 10 mM GABA or on synthetic dextrose ( SD ) medium supplemented with 10 mM ammonium sulphate ( [NH4]2SO4 ) as a sole nitrogen source [41 , 42] . At 37 °C , strains were grown on BHI medium or on SD medium supplemented with 10 mM ( NH4 ) 2SO4 . J774 murine macrophages ( 1 × 105 ) were seeded into each well of a 6-well microtitre tray containing one sterile coverslip and 2 mL of complete Dulbecco's Modified Eagle Medium ( complete DMEM: DMEM , 10% fetal bovine serum , 2 mM L-glutamine and penicillin-streptomycin ) . Macrophages were incubated at 37 °C for 24 h before activation with 0 . 1μg/mL−1 lipopolysaccharide ( LPS ) from E . coli ( Sigma ) . Macrophages were incubated a further 24 h at 37 °C and washed 3 times in phosphate buffered saline , and 2 mL of complete DMEM medium containing 1 × 106 conidia was added . A control lacking conidia was also performed . Macrophages were incubated for 2 h at 37 °C ( to allow conidia to be engulfed ) , washed once in phosphate buffered saline ( to remove free conidia ) , and incubated a further day at 37 °C . Macrophages were fixed in 4% paraformaldehyde and stained with 1 mg/mL−1 fluorescent brightener 28 ( calcofluor , CAL ) to observe fungal cell walls . The numbers of germinated conidia was measured microscopically by counting the numbers of germinated conidia ( conidia with a visible germ tube or yeast cells ) in a population of approximately 100 . Three independent experiments were performed . Mean and standard error of the mean values were calculated using GraphPad Prism3 . P . marneffei strains were grown on slides covered with a thin layer of solid medium , with one end resting in liquid medium [38] . Wild-type ( pakA+ ) , ΔpakA , ΔpakA pakA+ , and ΔpakA pakAH108G strains were grown on ANM medium supplemented with GABA at 25 °C for 2 or 4 d . At 37 °C , strains were grown on BHI medium for 4 d or in liquid BHI medium for 6 d . Immunofluorescence microscopy for examination of the actin cytoskeleton was performed using standard protocols [43] . Double staining was performed with the mouse C4 monoclonal anti-actin ( Chemicon International ) and rabbit anti-GFP polyclonal primary antibodies , as well as ALEXA 488 rabbit anti-mouse ( Molecular Probes ) and ALEXA 594 anti-rabbit ( Molecular Probes ) secondary antibodies . Single immunostaining controls and a minus primary antibody control were also performed . The gfp::pakA and gfp::pakAH108G strains were grown on agar-coated slides containing ANM plus GABA for 2 d or SD with 5 mM ammonium tartrate ( NH4T ) for 4 d at 37 °C . Slides were examined using differential interference contrast ( DIC ) and epifluorescence optics for GFP , antibody fluorescence , cell wall staining with fluorescent brightener 28 ( calcofluor , CAL ) , or nuclear staining with DAPI and viewed on a Reichart Jung Polyvar II microscope . Images were captured using a SPOT CCD camera ( Diagnostic Instruments ) and processed in Adobe Photoshop . Approximately 106 spores were inoculated into 300 μL of SD plus 10 mM ( NH4 ) 2SO4 and incubated for 8 , 12 , or 15 h at 25 °C or for 8 , 12 , 15 , or 20 h at 37 °C . The rates of germination were measured microscopically by counting the numbers of germinating conidia ( conidia with a visible germ tube ) in a population of 100 . Three independent experiments were performed . Mean and standard error of the mean values were calculated using GraphPad Prism3 . Two-level nested ANOVA was performed on the data for each time point at both 25 °C and 37 °C to test if germination rates differed significantly between genotypes and between transformants of the same genotype . ANOVA simultaneously tests two null hypotheses; there is no difference between the means of the data sets from all genotypes and there is no difference between the means of the data sets between transformants of the same genotype . The generation of two F-statistics and probability values allow rejection or acceptance of these null hypothesis at a 99% confidence . Values with an asterisk in Tables S1 and S2 showed significant differences between transformants of the same genotype . To investigate the differential regulation of conidial germination at 25 °C and 37 °C , wild-type and ΔpakA conidia were incubated in liquid medium at different temperatures ranging from 25 °C to 37 °C . The 20-h incubation was performed in a gradient thermocycler using 2 °C temperature increments from 25 °C to 37 °C . The media was then transferred to a microtitre tray and the rates of germination were measured microscopically by counting the numbers of germinating conidia ( conidia with a visible germ tube ) in a population of 100 . Three independent experiments were performed . Mean and standard error of the mean values were calculated using GraphPad Prism3 . | Many fungal infections are initiated by the entry of dormant fungal spores into their host . Once inside the host these dormant spores must reactivate ( germinate ) for the infection to proceed . Productive infections necessitate that the fungus grow and divide within the host , which makes understanding the mechanisms that control germination crucial to developing preventative or prophylactic treatments for fungal infections . The molecular mechanisms that control spore germination are poorly understood and studies of the opportunistic fungal pathogen Penicillium marneffei have shown that a group of highly conserved signalling and cell polarity factors , known as small GTPases , play important roles in germination and other aspects of morphogenesis . In this study we have shown that a downstream target of these small GTPases , a p21-activated kinase plays a crucial role in germination at the host temperature of 37 °C but not at 25 °C . This is the first component of germination , which shows temperature-dependent effects and provides insights into the different mechanisms used by fungal pathogens to infect their host or to grow saprophytically in non-host environments . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases",
"cell",
"biology",
"yeast",
"and",
"fungi",
"microbiology",
"genetics",
"and",
"genomics"
] | 2007 | A p21-Activated Kinase Is Required for Conidial Germination in Penicillium marneffei |
Arthropod-borne viruses ( arboviruses ) are among the most common agents of human febrile illness worldwide and the most important emerging pathogens , causing multiple notable epidemics of human disease over recent decades . Despite the public health relevance , little is know about the geographic distribution , relative impact , and risk factors for arbovirus infection in many regions of the world . Our objectives were to describe the arboviruses associated with acute undifferentiated febrile illness in participating clinics in four countries in South America and to provide detailed epidemiological analysis of arbovirus infection in Iquitos , Peru , where more extensive monitoring was conducted . A clinic-based syndromic surveillance system was implemented in 13 locations in Ecuador , Peru , Bolivia , and Paraguay . Serum samples and demographic information were collected from febrile participants reporting to local health clinics or hospitals . Acute-phase sera were tested for viral infection by immunofluorescence assay or RT-PCR , while acute- and convalescent-phase sera were tested for pathogen-specific IgM by ELISA . Between May 2000 and December 2007 , 20 , 880 participants were included in the study , with evidence for recent arbovirus infection detected for 6 , 793 ( 32 . 5% ) . Dengue viruses ( Flavivirus ) were the most common arbovirus infections , totaling 26 . 0% of febrile episodes , with DENV-3 as the most common serotype . Alphavirus ( Venezuelan equine encephalitis virus [VEEV] and Mayaro virus [MAYV] ) and Orthobunyavirus ( Oropouche virus [OROV] , Group C viruses , and Guaroa virus ) infections were both observed in approximately 3% of febrile episodes . In Iquitos , risk factors for VEEV and MAYV infection included being male and reporting to a rural ( vs urban ) clinic . In contrast , OROV infection was similar between sexes and type of clinic . Our data provide a better understanding of the geographic range of arboviruses in South America and highlight the diversity of pathogens in circulation . These arboviruses are currently significant causes of human illness in endemic regions but also have potential for further expansion . Our data provide a basis for analyzing changes in their ecology and epidemiology .
Over the past few decades there has been a global resurgence of arthropod-borne viral pathogens ( arboviruses ) worldwide [1] , [2] , particularly those transmitted by mosquitoes . Despite the public health relevance , the geographic range , relative impact , and epidemiologic characteristics associated with arbovirus infection are poorly described in many regions of the world . Arboviruses are a heterogeneous group , but those of medical relevance largely belong to a few virus genera , including Flavivirus , Alphavirus , and Orthobunyavirus . Prominent examples of emergent arboviruses include West Nile virus ( WNV; Flavivirus ) in North America , Japanese encephalitis virus ( JEV; Flavivirus ) in Asia , chikungunya virus ( CHIKV; Alphavirus ) in the Indian Ocean region and dengue viruses ( DENV; Flavivirus ) worldwide . One common feature shared by many emergent arboviruses is the capacity to expand host and geographical range , owing in part to the plasticity of the RNA genome [3] . Some arboviruses have evolved to exploit humans as their primary reservoir , while others rely on birds or peridomestic animals , with human infection resulting from spill-over from zoonotic replication cycles . Human exposure to sylvatic arbovirus cycles is likely to increase as activities continue to encroach on forested areas worldwide . Tropical areas in particular , with year-round hot and humid conditions , are well-suited for maintaining arboviruses with potential to emerge as significant human pathogens [4] . In the neotropics alone , greater than 145 distinct arbovirus species have been identified [4] , many of which have already been associated with human illness . One limitation of conducting surveillance for arboviral diseases is the generic nature of disease presentation . While severe disease can result , such as hemorrhagic manifestations ( DENV and yellow fever virus [YFV] ) or neurological disease ( WNV , JEV , and Venezuelan equine encephalitis virus [VEEV] ) , arbovirus infection typically results in mild to moderate febrile illness [2] , [5] , [6] . Particularly early in disease progression , patients commonly present with undifferentiated febrile illness [5] , [7] rendering clinical diagnosis unreliable [8] . In DENV-endemic areas , for example , diseases caused by co-circulating pathogens have been found to be often misclassified [8] , [9] . In light of the lack of distinct clinical presentation and the diversity of the etiologic agents , laboratory support has become a critical component of effective surveillance programs . The impact on human health in endemic regions and the potential for broader spread underscore the importance of improving understanding of arbovirus transmission patterns . Currently the epidemiological characteristics and geographic range for many endemic arboviruses in South America are poorly understood . To begin to address this gap , we established a laboratory-supported clinic-based study to identify the etiologic agents associated with undifferentiated febrile illness in sites in Peru , Ecuador , Bolivia , and Paraguay . Herein we describe the geographic distribution of distinct arboviruses and their relative contribution to human febrile illness in these study sites . In addition , we present the temporal trends and epidemiological characteristics associated with arbovirus infection in Iquitos , Peru , a site where more extensive monitoring was conducted .
In 1990 the U . S . Naval Medical Research Center Detachment ( NMRCD ) initiated a clinic-based surveillance program to determine the etiologies of febrile illness in Iquitos , Peru [10]–[12] . In 2000 NMRCD collaborated with local Ministries of Health to expand the surveillance program into other regions of Peru and South America , including sites in Ecuador , Bolivia , and Paraguay . In addition to Iquitos , in 2000 the study was implemented at regional sites in or near Piura , Cusco , Tumbes , and Yurimaguas , Peru , as well as Santa Cruz , Bolivia ( Figure 1; Table 1 ) . Additional sites were later added in Concepción , Magdalena , and Cochabamba ( Villa Tunari and Eterazama ) , Bolivia; Guayaquil , Ecuador; Asunción , Paraguay; and La Merced and Puerto Maldonado , Perú . Participants were recruited when reporting with acute febrile illness to public , private , or military health facilities in and around these regional centers . Details of the study sites are described in Table 1 . Study sites were selected based largely on locations in hot and humid climates conducive for arbovirus transmission , typically situated near or in tropical rainforest regions . Notable exceptions include Piura , Tumbes , and Cusco , which are located in coastal desert ( Piura and Tumbes ) or highlands ( Cusco; Figure 1 and Table 1 ) regions . It should be noted that the study staff in Cusco ( Hospital Regional ) on occasion attended to participants arriving from surrounding highlands rainforest regions . Study protocols ( NMRCD . 2000 . 0006 [Peru] , NMRCD . 2001 . 0002 [Ecuador] , NMRCD . 2000 . 0008 [Bolivia] , and NMRCD . 2005 . 0008 [Paraguay] ) were approved by the Naval Medical Research Center Institutional Review Board ( Bethesda , MD ) in compliance with all U . S . Federal regulations governing the protection of human subjects . In addition , the study protocols were reviewed and approved by health authorities in Peru ( Dirección General de Epidemiología ) , Bolivia ( Servicio Departamental de Salud , Santa Cruz and Colegio Medico de Santa Cruz ) , Ecuador ( Ministerio de Salud Publica , Comando Conjunto de la Fuerzas Armadas , and Escuela de Sanidad in Guayaquil ) and Paraguay ( Ministerio de Salud y Bienestar Social and Comité de Ética de Asociación de Rayos de Sol ) . Written consent was obtained from patients 18 years of age and older . For patients younger than 18 years , written consent was obtained from a parent or legal guardian . Additionally , written assent was obtained from patients between 8 and 17 years of age . Study subjects included patients 5 years of age or older who presented in outpatient clinics or hospitals with acute , undifferentiated , febrile illness ( greater than or equal to 38°C for 7 days duration or less ) along with one or more of the following symptoms: headache , muscle , ocular and/or joint pain , generalized fatigue , cough , nausea , vomiting , sore throat , rhinorrhea , difficulty breathing , diarrhea , jaundice , dizziness , disorientation , stiff neck , or bleeding manifestations . Children younger than five years of age were included if they presented with hemorrhagic manifestations indicative of dengue hemorrhagic fever ( DHF ) , including epistaxis , pleural effusion , platelets less than 100 , 000/ml , petechiae , or bloody stool or vomit . Exclusion criteria included fever in excess of seven days or an identifiable focus of infection , such as sinusitis , pneumonia , acute otitis media , or acute urinary tract infection . Demographic data , medical history , and clinical features for each patient were obtained using a standard questionnaire . In malaria-endemic regions if malaria was suspected , capillary blood from febrile patients was screened for Plasmodium spp . by clinic or hospital personnel according to routine diagnostic procedures at each site . Peripheral blood samples were screened by microscopic analysis of stained thick smear slides . In some sites , owing to the possibility of arbovirus co-infection , malaria-positive patients were subsequently invited to participate in the NMRCD study , with malaria results recorded along with symptoms and demographic information . During the acute phase of illness blood samples were obtained from each patient , and when possible , convalescent samples were obtained 10 days to 4 weeks later for serological studies . For patients older than 10 years of age , up to 15 mL of blood was collected , and for patients younger than 10 years of age , up to 7 mL of blood was collected . Trained phlebotomists collected blood samples via arm venipuncture using standard methods and universal precautions . Statistical analyses ( Chi-square , Fisher's exact test , and logistic regression ) were performed in R version 2 . 8 ( The R Foundation for Statistical Computing , Vienna , Austria ) [20] . The significance level was set at α = 0 . 05 .
DENV was the most common malaria co-infection , observed for 11 . 3% ( 66/584 ) of participants reported as malaria-positive , including 17 DENV-3 isolates and one DENV-1 isolate . The rate of DENV infection was significantly lower for malaria-positive participants than for malaria-negative participants ( 11 . 3% vs 32 . 9%; p<0 . 0001 ) . In contrast , VEEV infection was more common among malaria-positive participants ( 7 . 4% ) as compared to malaria-negative participants ( 2 . 7%; p<0 . 0001 ) . There were no significant differences between malaria thick smear-positive participants and thick smear-negative participants for the other arboviruses studied ( data not shown ) . DENV serotypes were the predominant arboviruses detected , accounting for 26 . 0% of febrile episodes analyzed ( Table 5 ) , based on virological ( 2 , 662 virus isolates or RT-PCR positives , with or without supporting serology ) and serological ( 1 , 058 IgM seroconversions and 1 , 700 participants with elevated DENV IgM , without accompanying positive results by virus isolation or RT-PCR ) evidence . Considerable YFV cross-reactivity was observed for DENV-positive samples . Based on the 2 , 662 cases with definitive DENV diagnosis ( IFA or RT-PCR confirmation in the acute sample ) , 847 ( 32 . 0% ) also had IgM reactive to YFV antigen in the acute or convalescent sample . DENV-3 was most commonly isolated serotype , accounting for 81 . 1% ( 2 , 159/2 , 662 ) of DENV isolates over the course of the study . In our study , DENV-3 was first detected in sites along the northern coast of Peru ( Piura and Tumbes ) in 2000 ( Figure 2 ) during a large outbreak of dengue fever in the region [21] , [22] , although DENV-1 and DENV-2 were the most commonly isolated serotypes during this outbreak . DENV-3 quickly became the dominant serotype in the northeastern rainforest ( Iquitos and Yurimaguas ) , with limited DENV-1 co-circulation in the region in subsequent years ( Figure 2 ) . Between 2002 and 2006 little DENV-2 transmission was observed until DENV-2 emerged in the study sites in Bolivia and southern Peru ( Puerto Maldonado ) in 2007 ( Figure 2 ) . DENV-4 was rarely detected during the study period , with only four isolates from study participants . However , more recently this situation has changed dramatically with the 2008 emergence of DENV-4 in northern Peru [23] . Overall , DENV infection was more common among female participants than male participants ( 28 . 1% vs 25 . 5%; p<0 . 0001 ) , with a statistically significant bias towards older participants ( 28 . 0% of participants 30 or older were DENV-positive as compared with 25 . 9% of participants younger than 30; p = 0 . 001 ) . However , the epidemiology of DENV infection varied by study site , particularly within Peru . The prevalence of DENV infection was higher among older participants in Puerto Maldonado ( p = 0 . 01 ) , La Merced ( p<0 . 0001 ) , Piura ( p<0 . 0001 ) , and Tumbes ( p = 0 . 015 ) , while no elevated DENV prevalence was observed for older participants in Yurimaguas and Iquitos . In addition , DENV infection was more common among female participants in Puerto Maldonado ( p<0 . 0001 ) , La Merced ( p<0 . 0001 ) , and Tumbes ( p = 0 . 028 ) , but no statistically significant difference was observed in Iquitos , Yurimaguas , or Piura . Other than DENV , the only flavivirus isolated during the course of the study was YFV , which was isolated from four participants ( Table 4 ) . In addition to the four isolates , serological evidence of recent YFV infection ( without evidence of DENV infection ) was detected in an additional 494 participants , including 143 who seroconverted between acute and convalescent samples . Overall , data on prior vaccination was available for 17 , 816 participants , 10 , 667 ( 59 . 9% ) of whom reported having received YF vaccination . YF vaccine coverage varied widely by study site , ranging from less than 10% in non-endemic sites along the northern coast of Peru ( Piura and Tumbes ) to 59% in Iquitos , 70% in Yurimaguas , and 77% or greater in Cochabamba ( Villa Tunari and Eterazama ) , Concepción , Junin , Magdalena , and Puerto Maldonado . Study participants with evidence of recent YFV infection based on IgM were significantly more likely than the rest of the overall study population to have reported receiving YF vaccination within the previous 6 months ( 30 . 0% vs . 6 . 7%; p<0 . 0001 ) . Recent alphavirus infection was detected for 3 . 1% ( n = 645 ) of febrile patients ( Table 5 ) , including 102 VEEV isolates and 40 MAYV isolates . RNA from a subset of VEEV isolates was extracted , reverse transcribed , amplified , and sequenced . All sequenced isolates were determined to belong to enzootic subtypes of the VEE complex , predominately ID Panama/Peru or Peru/Bolivia genotypes [12] , [14] , [24] although there was one ID Colombia/Venezuela genotype and one IIID subtype virus identified , both in Iquitos [12] . The majority of VEEV isolates and seroconversions ( 234/250; 93 . 6% ) were from patients in Iquitos , Puerto Maldonado , and Yurimaguas , Peru . In contrast , MAYV isolates were more prevalent in Bolivia and southeastern Peru ( Table 4 ) . Of all MAYV isolations , 57 . 5% ( 23/40 ) were from this region , despite representing only 19 . 8% of all participants in the study . EEEV was not isolated from any participant samples during the course of the study . A subset of participant samples were screened for EEEV-reactive IgM ( n = 3 , 014 ) , with serological evidence for EEEV infection in 22 cases ( 0 . 7% ) , including two seroconversions . Unlike the flaviviruses , little serologic cross-reaction ( or , alternatively , concurrent infection ) was observed among alphaviruses . For the 72 participants with the most well-defined VEEV infections ( IFA-positive , plus a convalescent sample available for testing ) , 7 ( 9 . 7% ) had IgM reactive to MAYV antigen in either the acute or convalescent sample . For the 24 cases where MAYV was isolated and a convalescent sample was available for testing , no cross-reactivity was observed in the VEEV IgM ELISA . Arboviruses belonging to the Orthobunyavirus genus of the Bunyaviridae family accounted for approximately 2 . 5% of all febrile cases ( Table 5 ) . In total there were 54 orthobunyavirus isolates , including 30 Group C viruses , 18 OROV isolates , and six GROV ( Table 4 ) . The Group C virus isolates were not definitively identified; however , based on serological techniques ( ELISA and PRNT ) , ten were antigenically related to CARV and six were antigenically related to MURV , while 14 could not be antigenically distinguished . Nearly all Simbu Group ( OROV ) and Group C virus isolates were collected from patients reporting to clinics in Iquitos , Madre de Dios , and Yurimaguas , Peru , while three out of six GROV isolates were obtained from patients in La Merced , Peru , in January and February of 2007 ( Table 4 ) . As with the alphaviruses , little serologic cross-reaction was observed within the Orthobunyavirus genus . For the 24 participants IFA-positive for a Group C virus and a convalescent sample available , only 2 had IgM reactive to OROV antigen in either the acute or convalescent sample ( 8 . 3% ) ; for the 12 OROV IFA-positive participants with a paired convalescent sample , no reactivity with CARV or MURV antigen was observed . The Iquitos health centers included in this study cover a geographically stratified area of the city and in 2007 represented nearly 20% ( 10 of 55 ) of civilian public health centers in the greater urban health network . Based on the populations assigned to each health center by the local ministry of health ( Dirección Regional de Salud -Loreto ) , in 2007 clinics included in this study were designated to serve approximately 43% of the population of the Iquitos region . Using the population numbers assigned to each health center by DIRESA-Loreto , we estimated incidence rates for the most common arboviral infections in Iquitos beginning with the first full year of the study ( Table 6 ) . Over the course of the study , there were 855 . 9 acute undifferentiated febrile episodes per 100 , 000 people per year , peaking during periods of highest dengue activity ( 2002 and 2004; Table 6 ) . DENV incidence rates varied greatly , peaking in 2002 with 554 . 0/100 , 000 following the introduction of DENV-3 and averaging 274 . 7/100 , 000 over the 7-year period . The average symptomatic incidence rates for other predominant arboviruses were 28 . 1/100 , 000 for VEEV , 8 . 5/100 , 000 for MAYV , 14 . 3/100 , 000 for OROV , and 14 . 2/100 , 000 for Group C viruses . Peak transmission rates were observed for these four viruses between 2004 and 2006 , including a previously-described outbreak of VEEV in 2006 [24] . It should be noted that the incidence rates above only reflect the participants enrolled in the study . Starting in 2006 , demographic data was collected for those who reported to Iquitos health centers and fulfilled the inclusion criteria ( acute undifferentiated febrile illness of fewer than 7 days in duration ) but declined participation in the febrile surveillance study . In 2006 and 2007 , 3 , 385 and 3 , 283 febrile patients , respectively , fitting the case definition were examined by study personnel , with 43 . 3% ( n = 1 , 433 ) and 36 . 5% ( n = 1 , 197 ) of patients agreeing to provide venous blood samples for the surveillance program . During these two years , 94 . 3% of febrile patients were first screened for malaria by thick smear , with 32 . 4% of those screened classified as positive . Malaria-negative patients were significantly more likely to accept participation in the surveillance study ( 51 . 3%; 2 , 185/4 , 256 ) than malaria-positive patients ( 9 . 5%; 193/2 , 036; p<0 . 001 ) . Children were significantly less likely to participate than adults ( 25 . 3% of eligible children vs . 44 . 0% of eligible adults chose to participate; p<0 . 001 ) . To begin to describe the epidemiology associated with these arboviruses in the Iquitos , demographic characteristics of participants with recent infection by the most common pathogens – DENV serotypes , VEEV , MAYV , OROV , and Group C viruses – were compared with the rest of the participating febrile population in Iquitos . YFV infection , as determined by positive IgM ELISA , was significantly associated with self-reported recent YF vaccination ( OR 2 . 30 , 95% CI 1 . 44—3 . 57 ) ; no similar association was observed for other arboviruses . Thus no further analyses were conducted for YFV IgM-positive participants . Overall , male participants were more common than female ( 51 . 4% vs . 48 . 6% ) , consistent with the population of Loreto Department as a whole ( 51 . 2% male; p = 0 . 77 ) [25] . The median age of study participants was 23 ( average 26 . 1 ) , with the highest percentage of participants between the ages of 15 and 29 . Both MAYV ( p = 0 . 003 ) and VEEV ( p = 0 . 009 ) infection were significantly more common among males , and this effect was only observed among the older age groups ( 15 years or older ) , suggesting an occupational exposure . A similar trend for higher prevalence of Alphavirus infection among males was observed in Yurimaguas and Puerto Maldonado , although these analyses were limited by small sample size . Group C virus infection was more common in males in these three sites , although the differences were only statistically significant in Puerto Maldonado ( p<0 . 01 ) . In Iquitos no significant differences were observed between sexes for DENV or OROV ( Table 7 ) . DENV infection was significantly more common among participants younger than 15 in Iquitos ( p = 0 . 005 ) ; however , this effect was only observed during the earlier years of the study ( 2002 and 2003 in particular ) . OROV infection in Iquitos was significantly more common among age groups 15 or older ( p = 0 . 007 ) . For the 30–44 year old age group , MAYV infection was significantly more common than for participants younger than 15 ( Table 7 ) . The health centers in the Iquitos area included in this study were predominantly public clinics and hospitals located within the urban area of the city ( n = 8 ) , although the study was also conducted in three military clinics located within the urban area and two public clinics located in rural zones between approximately five and ten kilometers outside the city limits . The majority of participants in the Iquitos area were recruited in the urban clinics ( 79 . 6% ) , while 10 . 6% and 9 . 8% were recruited at the two rural clinics and three military clinics , respectively . Using the different categories of clinics as a proxy for potential differences in arbovirus exposure , we compared the relative prevalence of arboviruses among those reporting to the urban , rural , or military clinics . DENV infection was far more common in participants reporting to the urban clinics than rural clinics , whereas VEEV ( p = 0 . 018 ) , MAYV ( p<0 . 001 ) , and Group C viruses ( p<0 . 001 ) were more common among those reporting to the rural clinics . For OROV infection there was no statistically significant differences among the types of health centers ( Table 7 ) . Participants were recruited year-round , with a peak in December that was largely due to DENV transmission ( Figure 3A ) . Transmission of the arboviruses peaked during different months of the year . Over the course of the study , DENV transmission was most common between October and December with lowest levels between June and August ( Figure 3A ) . Alphavirus transmission was highest between February and July ( Figure 3B ) , with reduced transmission during the second half of the year , while the highest percentage of Group C virus cases was observed between December and February ( Figure 3C ) .
Tropical areas , with year-round hot and humid conditions , are particularly well-suited for maintaining arboviruses with both current public health importance as well as the potential to emerge as significant human pathogens [4] . Therefore in this study we focused on arbovirus transmission in tropical regions of four countries in South America . Our data demonstrate that arboviruses are a common cause of human febrile illness in these sites in South America , accounting for greater than 30% of the febrile episodes analyzed . Importantly , arbovirus-associated disease was not restricted to DENV in most of the locations studied . The other arboviruses identified , including VEEV , MAYV , and OROV , in total were associated with approximately 8% of febrile episodes . Our study has provided source material for various phylogenetic analyses [12] , [14] , [19] , [23] , [26] and will provide important baseline data for monitoring changes in arbovirus ecology , epidemiology , and genetics . There were several significant limitations to our study . First , with the exception of Peru , the number of study sites in the other countries was quite limited . Even in Peru , it is unclear whether these results are indicative of arbovirus circulation in other regions of the country . Another shortcoming was the focus on arboviruses . While clearly these are important pathogens in the tropical rainforest regions , along the desert coast ( Piura ) and in the highlands ( Cusco ) , other types of pathogens will need to be given greater consideration . Another limitation of our study design is the passive surveillance strategy employed . Clinic-based surveillance is likely to significantly underestimate true arbovirus circulation , as those with milder disease manifestations are less likely to visit a health center . In studies of DENV transmission in Iquitos we have observed that incidence rates calculated from community-based active surveillance are several times higher than those calculated based on passive surveillance ( TJK , ACM , and BMF , unpublished results ) . Accordingly the incidence rates presented here should be interpreted carefully and considered a conservative estimate of the true number of febrile episodes caused by each virus . Another shortcoming of clinic-based surveillance is the difficulty of extrapolating the data to the entire population . As we show here in Iquitos , those who present to the health centers and those that are willing to participate in these studies are often not fully representative of the population at-large , which may lead to biases in age-dependent incidence rates . In addition , with the exception of Iquitos , we did not collect sufficient data from non-participants to fully contextualize these results . Our data suggest that malaria may contribute to approximately 30% of acute febrile illnesses in Iquitos , a figure that would not be apparent based solely on those who enrolled in the study . One advantage of clinic-based passive surveillance is expanded geographic coverage and more limited costs relative to other surveillance strategies , which is critically important when studying the relatively obscure arboviruses described here . Only through the large number of participants presented here were we able to detect sufficient cases of VEEV , MAYV , and OROV for further epidemiological analysis . Over the course of the study , DENV serotypes were by far the most common arboviruses associated with febrile disease , accounting for 26% of febrile participants . DENV serotypes have emerged dramatically in Latin America over the past decades , to the point that nearly a million cases of dengue fever are reported every year in Latin America , along with thousands of cases of more severe disease that may lead to hemorrhagic manifestations and death [27] , [28] . Here we demonstrate that DENV-3 ( previously identified as subtype III [26] ) was the predominant serotype in the region between 2001–2007 , although we also observed significant DENV-1 and DENV-2 transmission in certain regions . Not surprisingly DENV circulation was found to be more region-dependent than country-dependent . Specifically , Tumbes and Piura along the coast of northern Peru share common trends with Guayaquil in Ecuador ( Figure 2A ) , while DENV circulation in Puerto Maldonado in southern Peru is more closely tied to trends observed in Bolivia ( Figure 2C ) . More recently we observed that a genetically conserved strain of DENV-4 was identified in Ecuador ( 2006 ) , then coastal Peru ( 2007 ) , before spreading to the tropical rainforests of northeastern Peru ( Iquitos and Yurimaguas; 2008 ) [23] . As multiple serotypes have been circulating in the region severe disease resulting from heterologous secondary infection is increasingly likely to occur [27] , [29] . In this study we did not distinguish between primary and secondary infection , and thus further analysis will be needed to identify the genotypes [30] and prior DENV immune status associated with more severe disease outcomes in the region . Regardless , the data described here will provide a springboard for future studies of regional DENV maintenance and dispersion patterns [31] as well as analysis of genetic adaptation and selective pressures . Other than DENV , the only other flavivirus isolated was YFV . One well-documented hindrance to study flavivirus is the cross-reaction observed among even disparate species [32] , [33] . Similarly here we observed significant cross-reaction between DENV and YFV antigen in serum from patients with defined DENV infection , thus there is a possibility that some of the cases have been misclassified . For YFV , we only considered those instances where there was no DENV IgM detected . Furthermore , there was a strong correlation between participants reporting recent YF vaccination and having YFV-reactive IgM , suggesting that these results were not due to cross-reactivity with other flaviviruses circulating in South America . Low grade fever and headache are not uncommon outcomes within the two weeks following YF vaccination [34] , [35] , so it is possible that these cases are due to the vaccination . It should be noted , however , that flavivirus IgM can be long-lived [32] , and thus many of these febrile episodes classified as “YFV infection” may not represent the true etiologic agent . In addition , in our study we only rarely observed severe disease associated with YFV-reactive IgM , suggesting that these cases largely do not reflect natural infection and thus should be interpreted with caution . In addition to DENV and YFV , there are other flaviviruses circulating in the region that need to be considered , including WNV , Rocio virus ( ROCV ) , Ilheus virus ( ILHV ) , and St . Louis encephalitis virus ( SLEV ) . These flaviviruses have been isolated either from mosquitoes [36] , birds [37] , or mammals [38] , including humans [39] , [40] , in parts of South America . More closely related to our study areas , ILHV has been isolated from a febrile patient in Ecuador [41] , and ILHV and SLEV have been isolated from mosquitoes in Iquitos [36] , clearly demonstrating that these viruses are circulating near human populations in the region . None of these viruses were isolated from participants in our study , suggesting that human infection is uncommon . However , in a preliminary survey of a subset of our participants we have identified cases where ROCV , ILHV , SLEV , or WNV IgM was detected , with no reactivity with DENV or YFV antigen ( data not shown ) , with confirmation by virus neutralization assay , considered the most specific tool for flavivirus serology [42] . Overall , the cross-reactivity reported here and elsewhere [32] and the longevity of flavivirus IgM underscore the complications of flavivirus serodiagnosis , which represents a great hindrance for epidemiological surveillance . The most common Alphavirus species identified were VEEV and MAYV . Scant evidence for human infection with EEEV was identified , consistent with previous reports [43] , despite evidence of EEEV circulation in mosquitoes near Iquitos [36] , [44] , for example . In light of the recent emergence of another alphavirus , CHIKV , in the Indian Ocean region , VEEV and MAYV represent interesting cases to consider with regards to potential for urban emergence . In laboratory studies Aedes spp . , the primary vectors for DENV , have been shown to be a competent vector for VEEV [45] and MAYV [46] . Even without adapting to human-Aedes-human cycles , epizootic VEEV subtypes have been associated with large outbreaks of human disease across South America [6] . As recently as 1995 a VEEV outbreak was responsible for nearly 100 , 000 febrile cases in Venezuela and Colombia [47] , [48] . While the VEEV strains isolated in our study all belong to enzootic genotypes of the virus complex [12] , [14] , [24] , genetic studies have demonstrated that enzootic and epizootic subtypes are closely related . A modest number of amino acid changes can alter the viral phenotype dramatically , converting an enzootic strain to an epizootic strain [49]–[52] . Similarly , amino acid variants in the CHIKV E1 protein have been associated with increased epidemic potential [53]–[55] . Several other factors further suggest that potential for neotropical alphavirus emergence is high . In the Iquitos area , while we found that VEEV was more commonly associated with rural clinics ( Table 7 ) , many of the participants with confirmed VEEV infection lived within the city and did not report leaving the urban area during the month prior to the febrile illness [24] ( data not shown ) . This data is corroborated by a previous study of healthy participants , in which we found that nearly 25% of the urban population carries VEEV-neutralizing antibodies [24] . In addition , based on data collected through this program the geographic range of MAYV and VEEV is wider than had been previously demonstrated , extending to southern Peru and Bolivia [14] , [19] . Taken together these factors suggest that the potential establishment of the neotropical alphaviruses as urban pathogens should be closely monitored . In addition to the flaviviruses and alphaviruses , orthobunyaviruses were significant sources of febrile illness in the study , accounting for 2 . 5% of febrile episodes analyzed . While all orthobunyavirus isolates came from patients in Peruvian rainforest sites during the course of this study , we did find serological evidence for OROV and Group C viruses in Ecuador and Bolivia . More recently ( 2008 ) we have definitively identified Group C viruses in Bolivia , isolated from two participants in the Cochabamba region ( data not shown ) . Like VEEV and MAYV , OROV is an interesting case study with regards to potential for broader emergence . OROV has been associated with numerous outbreaks in Brazil , infecting approximately 500 , 000 people in South America over the past 45 years [56]–[58] . Two distinct transmission cycles have been proposed , a poorly-defined sylvatic cycle and an urban cycle where OROV is transmitted among humans by the biting midge Culicoides paraensis [58] , [59] . In Iquitos , we found that unlike the Group C viruses , VEEV , and MAYV , evidence of recent OROV infection ( based on both IgM and virus isolation data ) showed no significant bias towards rural clinics , suggesting similar transmission levels between urban and rural sites , consistent with results from an earlier survey of healthy participants in the region [10] . This pattern may reflect a peri-urban transmission cycle , as the majority of the OROV isolates were detected in both the rural sites and an urban site located towards the periphery of the city in 2005 and 2006 during a period of markedly increased transmission ( Table 6 ) . OROV isolates from previous Iquitos studies ( prior to 1998 ) were found to belong to lineage II , similar to strains associated with Brazilian OROV outbreaks [56] , [60] . Future sequence analysis of the more recent isolates described in this current study from Iquitos , Yurimaguas , and Puerto Maldonado , will provide a more complete description of OROV geographic and temporal genetic variability . Considering the association of arboviral pathogens with human disease and the potential for wider-scale emergence , disease surveillance is an integral component of public health planning , disease control , and analysis of potential intervention strategies . Unfortunately , for the arboviruses described here syndromic surveillance is complicated by the lack of pathogen-specific signs and symptoms [5] , particularly early in disease progression . As with other reports [7] , [9] , our study underscores the need for laboratory-based support of febrile surveillance studies . Even within our study other pathogens clearly need to be considered , as the majority of febrile episodes in this study were not associated with an arboviral etiology . In Iquitos past studies have linked both Leptospira spp . and Rickettsia spp . with a significant percentage of febrile illnesses [61]–[63] . To-date , solid data are lacking for the other study sites included in this study , although our preliminary results suggest that Leptospira spp . and Rickettsia spp . are common human pathogens in these locations as well ( TJK , unpublished results ) . Admittedly our studies provide little in the way of guidance for patient care but do point toward the need for the development of pharmaceutical therapies for the treatment of a variety of viral infections . In addition the development of rapid and inexpensive diagnostic tools should be given high research priority , particularly to distinguish arbovirus infection from other pathogens where effective and inexpensive pharmaceutical treatment is already available , such as for Rickettsia spp . and Leptospira spp . | Over recent decades , the variety and quantity of diseases caused by viruses transmitted to humans by mosquitoes and other arthropods ( also known as arboviruses ) have increased around the world . One difficulty in studying these diseases is the fact that the symptoms are often non-descript , with patients reporting such symptoms as low-grade fever and headache . Our goal in this study was to use laboratory tests to determine the causes of such non-descript illnesses in sites in four countries in South America , focusing on arboviruses . We established a surveillance network in 13 locations in Ecuador , Peru , Bolivia , and Paraguay , where patient samples were collected and then sent to a central laboratory for testing . Between May 2000 and December 2007 , blood serum samples were collected from more than 20 , 000 participants with fever , and recent arbovirus infection was detected for nearly one third of them . The most common viruses were dengue viruses ( genera Flavivirus ) . We also detected infection by viruses from other genera , including Alphavirus and Orthobunyavirus . This data is important for understanding how such viruses might emerge as significant human pathogens . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"virology/emerging",
"viral",
"diseases"
] | 2010 | Arboviral Etiologies of Acute Febrile Illnesses in Western South America, 2000–2007 |
It is important that bacterium can coordinately deliver several effectors into host cells to disturb the cellular progress during infection , however , the precise role of effectors in host cell cytosol remains to be resolved . In this study , we identified a new bacterial virulence effector from pathogenic Edwardsiella piscicida , which presents conserved crystal structure to thioredoxin family members and is defined as a thioredoxin-like protein ( Trxlp ) . Unlike the classical bacterial thioredoxins , Trxlp can be translocated into host cells , mimicking endogenous thioredoxin to abrogate ASK1 homophilic interaction and phosphorylation , then suppressing the phosphorylation of downstream Erk1/2- and p38-MAPK signaling cascades . Moreover , Trxlp-mediated inhibition of ASK1-Erk/p38-MAPK axis promotes the pathogenesis of E . piscicida in zebrafish larvae infection model . Taken together , these data provide insights into the mechanism underlying the bacterial thioredoxin as a virulence effector in downmodulating the innate immune responses during E . piscicida infection .
Thioredoxins ( Trxs ) are small redox-active molecules ubiquitously expressed in all taxa , from bacteria to mammals , containing a conserved redox catalytic CXXC ( -Cys-X-X-Cys- ) -motif that links the second β-strand to the second α-helix [1] . In mammalian cells , interactions between cytosolic TRX1 and apoptosis signal-regulating kinase 1 ( ASK1 ) suppress the activation of c-Jun N-terminal kinase ( JNK ) - , Erk1/2- , and p38-MAPK signaling cascades in response to various stress stimuli and activate a number of transcription factors that regulate various aspects of cell growth and survival [2–8] . Mitogen-activated protein kinase ( MAPK ) signal transduction , which involves sequential activation and amplification of downstream kinases , is a high-value target of bacterial pathogens during infection process [9] . Bacteria have developed several strategies to target MAPK pathway in order to subvert their functions , one of which is that bacterial virulence factors operate as mimics of host proteins [10–11] . Anthrax toxin ( lethal factor , LF ) , produced by the bacterium Bacillus anthracis , was proved directly inhibiting MAPKs by cleavage of the amino terminus of MAPKK1 and MAPKK2 [12] . Since then , YopJ from Yersinia pestis [13–14] and AvrA from Salmonella [15] was proved function as acetyl transferases that covalently modify key serine and threonine residues of MAPKs , regulating the transcription of pro-survival genes during infection . OspF from Shigella flexneri , which is homologous to the Salmonella SpvC and Pseudomonas syringae HopAI1 , possesses the phosphor-threonine lyase activity , and irreversible dephosphorylate MAPKs by covalent modification and inhibit the inflammatory responses [16–17] . Edwardsiella piscicida , previously named Edwardsiella tarda , is an intracellular bacterium with broad cellular tropism , which is a pathogen primarily for fish [18–19] . Type III secretion system ( T3SS ) and type VI secretion system ( T6SS ) have been identified in this pathogen [20–23] . Moreover , E . piscicida EIB202 activates NLRC4 and NLRP3 inflammasomes via T3SS and inhibits the NLRP3 inflammasome via the T6SS effector EvpP [22] . Recently , we revealed that the wild-type E . piscicida ( EIB202 ) replicates and induces pyroptosis in macrophages [24] . The macrophage-released E . piscicida population exhibits enhanced infectivity both in vitro and in vivo [24] , and displays a reprogrammed transcriptional profile characterized by the upregulation of T3SS/T6SS-related genes as well as some uncharacterized genes [20–21 , 23–24] . Thus , we hypothesize that the infection-induced E . piscicida genes might play important roles during infection . In this study , we found that one of the most significant infection-induced E . piscicida genes , trxlp , could be secreted and translocated into the cytosol of host cells upon infection , showing a conserved crystal structure to the host thioredoxin protein . Furthermore , Trxlp could mimic endogenous TRX1 to directly target the TRX-binding domain of ASK1 ( ASK1-TBD ) and inhibit its activation , subsequently suppressing Erk/p38-MAPK signaling , correlating to limit bacterial virulence and replication in vivo . Collectively , this study advances our understanding of bacterial thioredoxin as a virulence effector that mimics the host endogenous protein in manipulating innate immunity .
Based on the comparison of the global gene transcriptional profiles between macrophage-released and DMEM-cultured EIB202 [24] , we annotated 10 most highly upregulated genes of EIB202 after macrophage infection ( S1 Table ) . The gene of ETAE_2186 was among the top 3 upregulated genes and was annotated as a hypothetical thioredoxin ( S1 Table ) ; here , we named it Trx-like protein ( Trxlp ) . Notably , during EIB202 infection in macrophages , trxlp was dramatically upregulated , while the other 10 Trx antioxidant family proteins in this strain were not significantly induced during infection ( S1 Fig ) . E . piscicida possesses multiple secretion systems to deliver virulence factors during infection [20–23 , 25] . To investigate whether Trxlp is a secreted protein , we assessed the production of Trxlp-HA in EIB202 strains grown in DMEM . The robust secretion of Trxlp , but not classical Trx1 or Trx2 , was detected in the bacterial supernatants ( Fig 1A ) . Importantly , comparable production of Trx1 , Trx2 , and Trxlp-HA was observed in the pellets of EIB202 ( Fig 1A ) . In addition , Trxlp was secreted in T3SS-deficient or T6SS-deficient EIB202 at a comparable level to that in wild-type strains ( Fig 1B ) . Moreover , Trxlp was not detected in the fraction of the outer membrane vesicles ( OMVs ) ( Fig 1C ) , which were reported to be vehicles delivering bioactive proteins , toxins , and virulence factors [26 , 27] . Thus , these results indicate that the Trxlp secretion was independent of T3SS , T6SS , or OMV pathways . We next investigated the intracellular localization of Trxlp during infection . The subcellular fractionation of HeLa cells infected with EIB202 revealed that Trxlp is translocated into HeLa cells and localized in the cytosolic fraction , but not in the membrane fraction ( Fig 1D ) . Moreover , immunofluorescence microscopy revealed that Trxlp , but not Trx1 and Trx2 , localized into the cytosol of HeLa cells ( S2 Fig ) . Furthermore , infection of HeLa cells revealed that Trxlp , but not Trx1 or Trx2 , could be translocated into host cells through the β-lactamase reporter system , and the translocation of Trxlp was independent of T3SS or T6SS ( Fig 1E ) . Thus , our results suggest that Trxlp is a novel E . piscicida virulence effector that can be translocated into the cytosol of host cells during infection . To better characterization the importance of this virulence effector , we solved the crystal structure of Trxlp by molecular replacement using Thermus thermophilus Trx ( PDB:2YZU ) as the initial model . The structure was refined to an R-value of 17 . 7% ( Rfree = 21 . 5% ) with good geometry at a resolution of 1 . 98 Å ( S2 Table ) . Trxlp displays a similar structure to that of the canonical Trx fold , consisting of a β1-α1-β2-α2-β3-α3-β4-β5-α4 topology ( Fig 2A ) . The structure obtained here is in reduced form , as evidence by the 3 . 3 Å distance between 2 sulfur atoms of 2 cysteines in the CXXC motif ( Fig 2B ) . Trxlp contains a central β-sheet composed of 3 parallel ( β1 , β2 , β3 ) and 2 antiparallel strands ( β4 , β5 ) . This central β-sheet is sandwiched by 2 layers of helices: a bottom layer with 2 helices ( α1 , α3 ) and a top layer with the other 2 helices ( α2 , α4 ) . Trxlp can be superimposed to Trxs from other species , with root-main-square deviation ( rmsd ) values ranging from 1 . 54 to 1 . 74 Å for around 100 equivalent Cα pairs ( Figs 2C and S3 ) . Minor conformational differences among these structures were observed in flexible loop regions connecting α helices and β strands . Sequence-based homology searches and motif and PHYRE2 fold recognition analyses revealed that Trxlp contains a CXXC motif comparable to that of classical Trxs in other bacteria ( Fig 2B and 2D ) . Unlike Trx2 identified in EIB202 , both Trx1 and Trxlp contain only one CXXC-motif site and lack a mitochondrial-targeting sequence ( S4A Fig ) . Collectively , Trxlp was identified as a Trx family protein with a conserved redox catalytic CXXC motif . Thioredoxin is a ubiquitous thiol oxidoreductase that regulates the cellular redox status [7 , 8] . Thus , we analyzed the enzymatic function of Trxlp . Consistent with the number of CXXC motifs , the incubation of purified Trx2 resulted in a further reduction in the artificial disulfide substrate DTNB [28] compared to incubation with purified Trx1 or Trxlp ( S4B Fig ) . Moreover , the reduction of insulin by dithiothreitol ( DTT ) [29 , 30] at pH 7 . 0 was assessed in the absence or presence of EIB202 Trx1 , Trx2 , and Trxlp . Notably , both Trx1 and Trx2 catalyzed the reduction of insulin by DTT , as quantified by the onset of aggregation , while Trxlp was much less active than the classical reducing Trxs in EIB202 ( S4C Fig ) . Taken together , unlike the classical Trxs , Trxlp might not exhibit robust redox activity . Thus , this novel Trx family protein and its role during E . piscicida infection should be characterized further . TRX1 , an endogenous ubiquitous oxidoreductase , is a physiological inhibitor of ASK1 via interactions with its N-terminal region , termed the Trx-binding domain in human cells ( ASK1-TBD ) [4 , 31] . Trxlp shares a conserved structure and critical WCXXC-motif site not only with bacterial Trxs , but also with mammalian Trxs ( Figs 2C , 2D and S3 ) . Thus , we hypothesize that Trxlp can mimic endogenous TRX1 when it is translocated into the cytosol of host cells during infection . To determine whether Trxlp can interact with ASK1-TBD , we analyzed the immunoprecipitation of ASK1-TBD followed by immunoblotting to detect Trxlp , which showed that ASK1 associates with Trxlp via the Trx-binding domain ( Fig 3A ) . Moreover , an in vitro pulldown assay also revealed an association between Trxlp and ASK1-TBD ( Fig 3B ) . Notably , either mutation in Trxlp ( W50F , C51/53S , FSXXS ) or mutation in ASK1-TBD ( C250S ) significantly impaired the interactions between them ( Fig 3C and 3D ) . Thus , consistent with previous results regarding endogenous TRX1 binding with ASK1[4 , 31] , we identified , for the first time , a bacterial Trx family protein that can bind to ASK1-TBD via the conserved redox catalytic WCXXC motif . Since the N-terminal Trx-binding domain of ASK1 is necessary and sufficient for its association with Trx , which inhibits ASK1 activity by disrupting N-terminal coiled-coil ( NCC ) domain homophilic interactions [32] , it is interesting to test whether Trxlp could inhibit the homophilic interaction via the NCC domain of ASK1 ( ASK1-N ) . Thus , we examined the effect of Trxlp and human TRX1 on the homophilic interaction of ASK1-N by cotransfection analysis . The association of ASK1-N-HA with ASK1-N-Flag was inhibited by both human TRX1 and Trxlp in vitro ( Fig 3E ) . These findings suggest that the NCC domain-mediated homophilic interaction of ASK1 is suppressed by the association with Trxlp in vitro . The phosphorylation of the conserved threonine residue at the activation loop is essential for the kinase activity of human ASK1 [33] . In this study , we further analyzed the effects of Trxlp on the kinase activity of ASK1 . We detected robust phosphorylation of ASK1 in the presence of TNF-α , but this phosphorylation was significantly reduced in the presence of wild-type Trxlp or human TRX1 as a control ( Fig 3F ) . However , the mutant Trxlp ( FSXXS ) did not abrogate the TNF-α-induced ASK1 phosphorylation ( Fig 3G ) . Taken together , our results reveal that Trxlp can mimic endogenous TRX1 to inhibit the phosphorylation of ASK1 . During HeLa cells infection , Δtrxlp triggered robust ASK1 phosphorylation compared to EIB202 , and Trxlp complementation neutralized ASK1 phosphorylation , suggesting that Trxlp critically participates in regulating ASK1 activation ( Fig 4A ) . Since ASK1 is an upstream signaling partner of the MAPKKK family and its phosphorylation induces the activation of the MAPK signaling cascade in response to various stress stimuli [3–5 , 34] , it is interesting to examine whether Trxlp regulates the activation of the MAPKs or NF-κB pathways during E . piscicida infection . Interestingly , significantly increased phosphorylation levels of Erk1/2 and p38 were induced in HeLa cells infected by Δtrxlp than by the isogenic wild-type strain , or Δtrx1 and Δtrx2 ( S5A Fig ) ; however , no detectable difference in JNK phosphorylation and IκBα degradation was observed between them ( S5A Fig ) . In addition , we demonstrated that the phosphorylation of the Erk1/2- and p38 -MAPK pathways was induced by TNF-α in untransfected HEK293T cells , but it was dramatically reduced in cells expressing either human TRX1 or Trxlp ( S5B Fig ) ; meanwhile , the suppression of Erk1/2 and p38 phosphorylation was abrogated in cells expressing the mutant Trxlp ( FSXXS ) ( S5C Fig ) . Thus , these results suggest that Trxlp can suppress the phosphorylation of Erk1/2- and p38-MAPK signaling via the conserved redox catalytic WCXXC motif . To further validate the role of ASK1 in E . piscicida-induced MAPK pathway activation , the CRISPR/Cas9 genome-editing tool was applied to knockout ( KO ) ASK1 in HeLa cells using ask1-specific guide RNA . The Trxlp-suppressed Erk-1/2 activation effect was abrogated in ASK1-KO HeLa cells , while the MEK1/2 activation was not affected during E . piscicida strains infection ( Fig 4B ) . Moreover , the Trxlp-suppressed p38α activation effect was also abrogated in ASK1-KO HeLa cells , but the activation of MKK3/6 was not regulated by Trxlp ( Fig 4C ) . Simultaneously , consistent with above results , the activation of JNK , or its upstream kinase MKK7 was also not regulated by Trxlp ( S5D Fig ) . Collectively , these results suggest that Trxlp could inhibit the activation of ASK1 and thus limit Erk1/2- and p38-MAPK pathways through a MAPKK-independent mechanism during infection . Given that E . piscicida is a broad-range intracellular pathogen affecting from fish to mammals [18] , we subsequently investigated the role of Trxlp using a zebrafish fibroblasts ( ZF4 ) . Upon infection , ASK1 phosphorylation was enhanced in cells infected with Δtrxlp compared to the isogenic wild-type strain , but it was reduced to normal level when ZF4 cells were infected with a trxlp-complemented strain ( Fig 4D ) . Consistent with the results obtained using mammalian cells , Erk1/2 and p38 phosphorylation levels were also enhanced in ZF4 cells infected with Δtrxlp , but not in cells infected with trxlp-complemented strain ( Fig 4D ) . Furthermore , comparable phosphorylation levels of JNK and MAP2Ks , including MEK1/2 , MKK3/6 , MKK4 , and MKK7 , were observed between wild-type and Δtrxlp infected ZF4 cells ( S6A Fig ) . It is known that p38-MAPK or Erk1/2-MAPK potently control the production of many pro- or anti-inflammatory cytokines , which are critical for host immunity [9 , 35 , 36] . Thus , we further analyzed the regulation of cytokine production during E . piscicida infection . Δtrxlp induced the greatly increased transcription of TNF-α and IL-10 than wild type strain in ZF4 cells , which was counteracted by Trxlp complementation ( S6B and S6C Fig ) . Simultaneously , ELISA assay demonstrated that the production of TNF-α was significantly induced in cells infected with the Trxlp mutant strain ( S6G Fig ) . However , the transcription levels of IL-6 , cxcl8 , and IFN-γ were not regulated by Trxlp during E . piscicida infection ( S6D to S6F Fig ) . Taken together , these observations demonstrate a key role of Trxlp in suppression of ASK1-Erk/p38-MAPK signaling and thereby regulation of inflammatory cytokines expression during E . piscicida infection . Given the similarity between zebrafish MAPK pathways and those of mammals [37 , 38] , to evaluate the function of Trxlp in the regulation of ASK1-MAPK signaling in vivo , we developed a microinjection infection model [38] using 3 days post fertilized ( dpf ) zebrafish larvae for analyzing the pathogenesis of E . piscicida ( Fig 5A ) . We found that zebrafish larvae were more susceptible to EIB202 or trxlp-complemented strain than to Δtrxlp during infection ( Fig 5B ) , consistent with the reduced pathogen loads in Δtrxlp-infected zebrafish larvae ( Fig 5C ) . In addition , we found that infection with EIB202 induced the expression of TNF-α and IL-10 transcripts , which was further enhanced in zebrafish larvae infected with Δtrxlp ( Fig 5D ) . The enhancement of cytokine expression observed with Δtrxlp was abrogated when zebrafish larvae were infected with the trxlp-complemented strain ( Fig 5D ) . However , the transcript levels of IL-6 , cxcl8 , and IFN-γ were not regulated by Trxlp during E . piscicida infection ( S7B to S7D Fig ) . These results indicate that the novel virulence effector Trxlp facilitates bacterial survival and virulence in vivo . To analyze the involvement of ASK1 in promoting immune defense against E . piscicida in vivo , the morpholino ( MO ) oligonucleotide was designed to block ask1 translation and injected into embryos at the one-cell stage . Immunoblotting showed that this MO effectively knocked down ask1 for up to 7 dpf ( S7A Fig ) . Following infection with 50 CFUs of EIB202 or Δtrxlp , ask1-MO larvae succumbed more rapidly than the control-MO larvae ( Fig 5B ) . Consistent with a role of ASK1 in maintaining homeostasis , ask1-MO larvae had significantly higher pathogen loads during either EIB202 or Δtrxlp infection than in control-MO larvae ( Fig 5C and S3 Table ) . Furthermore , ask1-MO larvae exhibited significantly reduced TNF-α and IL-10 transcript levels during either EIB202 or Δtrxlp infection ( Fig 5D ) , while the transcript levels of IL-6 , cxcl8 , and IFN-γ were not affected compared with infected control-MO larvae ( S7B to S7D Fig ) . Taken together , our results suggest that the ASK1 might be a host target of the E . piscicida virulence effector Trxlp during infection in vivo , and the activation of ASK1-MAPK signaling cascades plays critical role in innate immunity ( Fig 5E ) .
Trx was originally considered an important conserved family for protection against ROS by reducing peroxides to harmless products [7–8] . The antioxidant defense system of microorganisms comprises various conserved antioxidant molecules [8] , but little is known about the specific roles of these molecules during infection . In this study , we analyzed all annotated Trx family proteins of E . piscicida , and found that only Trxlp was significantly upregulated when compared with the levels of classical bacterial Trx family proteins during infection ( Figs 1A and S1 ) . However , this Trxlp showed significantly lower redox activity than the classical reducing Trxs in EIB202 ( S3 Fig ) . This is the first report that a unique bacterial thioredoxin was utilized as a virulence effector to interfere with host antibacterial signaling , which expands our understanding of the bacterial Trx family proteins that not only catalyze protein disulfide reductase , but also function as virulence effectors during infection . In mammals , endogenous Trxs maintain the cellular redox state and regulate cell proliferation by acting as electron donors of ribonucleotide reductase [7–8] . Previous studies have shown that endogenous TRX1 is a negative regulator of ASK1 and constantly forms an inactive complex with ASK1 by associating with the N-terminal regulatory domain and inhibiting homophilic interactions with ASK1 [4 , 31 , 32] . Moreover , the ring finger domains of TRAF2 and TRAF6 downstream are required to accelerate the N-terminal homophilic interaction of ASK1 [32] , and the deregulation of ASK1 affects cell fate , such as survival and apoptosis , which is required for mammalian innate immunity [3 , 39] . Interestingly , we identified a bacterial virulence effector , Trxlp , which could mimic endogenous TRX1 to bind with the ASK1 and block its homophilic interactions , however , the precise mechanism of bacterial Trxs in manipulating host innate immune signaling during infection remains an open question . Previous studies have shown that ASK1 is activated in response to a variety of stress-related stimuli via distinct mechanisms and activates MKK4 and MKK3 , which in turn activate JNK and p38 [3] . In our study , either in mammalian cells or fish cells infected with E . piscicida , Trxlp could regulate the phosphorylation of ASK1 . However , quite different from previous observations , we found that the bacterial infection-engaged inhibition of ASK1 was responsible for regulating Erk1/2- and p38-MAPKs activation , but not JNK-MAPK signaling ( Figs 4 and S5 ) . Although , this was consistent with a previous study that observed the down-regulation of Trx induced ROS-mediated ASK1-Erk/p38-MAPK activation in human promonocyte cells during Japanese encephalitis virus infection [5] , we still expecting that multiple effectors might alter the JNK-MAPK pathway adversely during infection . Besides , another interesting issue is the activation of MAP2K pathways were not affected during E . piscicida infection , which was quite different from the classical activation of ASK1-MAP2K-MAPK signaling cascades under stress-related stimulus [3] . This discrepancy suggests a possibility that a MAP2K signaling independent pathway might be triggered by ASK1 to activate p38- and Erk1/2-MAPKs during infection . In E . piscicida , multiple effectors , especially those using T3SS and T6SS systems , cooperate or feedback with each other to directly mimic , intercept , or modify the function of key host factors engaged in a wide range of cellular processes , including innate immune signaling , cytoskeletal dynamics , membrane trafficking , phosphoinositide lipid metabolism , and cell signaling , which finally influence bacterial dissemination and survival [24] . EseH inhibited phosphorylation of Erk1/2 , p38α and JNK MAPK pathways in host cells , but had no effect on the NF-kB pathway [40] . EvpP significantly suppressed JNK activation , thus impairing oligomerization of the inflammasome adaptor ASC [22] . EseK also can inhibit MAPK phosphorylation and promotes bacterial colonization in zebrafish larvae [41] . Here , we present a comprehensive functional interpretation for the newly-identified virulence effector mimicking host Trx to regulate host ASK1-Erk/p38 MAPKs axis and promote bacterial infection in vivo ( Fig 6 ) . However , many details upstream of its intracellular behaviors , including how the bacterium responds to infectious signals to upregulate Trxlp transcription and how the secretion and translocation of Trxlp are coordinated during infection remain unknown . Thus , supplementary with our previous identified effectors in regulating MAPK pathways , our study clarified the roles of Trxlp in inhibited Erk1/2 , p38α-MAPK pathways in host cells , and the dynamic or redundant roles of these effectors in regulating MAPK signaling during E . piscicida infection to achieve its infectious goal remains to be clarified . Taken together , our results provide the first comprehensive , functional analysis of the E . piscicida Trx family protein Trxlp , a novel virulence effector that mimics endogenous TRX1 to target ASK1 and suppress its activation , thereby inhibiting the phosphorylation of Erk1/2- and p38-MAPKs and disrupting the expression of inflammatory cytokines and diminishing the host antibacterial defense . These findings provide insight into the mechanisms underlying the regulation of ASK1 , suggest that the development of drugs targeting ASK1 may be useful for the treatment of bacterial infectious diseases , and advance our knowledge of the general biology of pathogen–host interactions .
Animal experiments were conducted according to the Guide for the Care and Use of Medical Laboratory Animals ( Ministry of Health , People’s Republic of China ) and ethically approved by The Laboratory Animal Ethical Committee of East China University of Science and Technology ( Protocol #2006272 ) . All the infection experiments were conducted as a completely randomized design , and the analyses were performed in a blinded manner . The bacterial strains used in this study are described in S4 Table . Wild-type E . piscicida EIB202 ( CCTCC M208068 ) and indicated mutants were grown in tryptic soy broth ( TSB; BD Biosciences ) , tryptic soy agar ( TSA ) or Dulbecco’s modified eagle medium ( DMEM; Invitrogen ) at 30°C . Escherichia coli was cultured in Luria-Bertani ( LB; BD Biosciences ) broth or agar at 37°C . Antibiotics were added to the media at the following concentrations: ampicillin ( Amp ) , 100 μg/ml; kanamycin ( Km ) , 50 μg/ml; colistin ( Col ) , 16 . 7 μg/ml . HEK293T ( ATCC CRL-11268 ) , J774A . 1 ( ATCC TIB-67 ) and HeLa cells ( ATCC CCL-2 ) were grown at 37°C in DMEM supplemented with 10% FBS and under a 5% ( vol/vol ) CO2 atmosphere . ZF4 cells ( ATCC CRL-2050 ) , established from 1-day-old zebrafish embryos , were grown at 28°C in DMEM/F12 medium supplemented with 10% FBS and under a 5% ( vol/vol ) CO2 atmosphere . To detect the secretion of Trxs protein , wild-type , ΔT3SS and ΔT6SS E . piscicida expressing Trxs-HA protein were constructed , respectively . The Trx1 , Trx2 and Trxlp sequences were amplified from E . piscicida genome and the constructed plasmids containing arabinose operon pUTt-pBAD-trx1 , trx2 , trxlp-HA were electroporated into indicated strains [42] . To construct the plasmids expressing Trxlp , hTRX1 , N-terminal fragments of human ASK1-thioredoxin binding domain ( ASK1-TBD , DNA encoding residues 88–302 ) and N-terminal coiled-coil domain of ASK1 ( ASK1-N , DNA encoding residues 400 ) in eukaryotic cells , the plasmid pCDH-CMV-MCS-EF1-Puro ( CD510B-1 ) was linearized with Xba I and EcoR I , and then the PCR products of genes containing compatible ends were inserted into the linearized pCDH plasmid using one step cloning kit ( Vanzyme ) . To construct the Δtrx1 , Δtrx2 and Δtrxlp E . piscicida , an in-frame deletion mutation of trxs was generated by sacB-based allelic exchange as described [43 , 44] . For example , the upstream and downstream fragments of trxs were fused by overlapping PCR . Primer pairs deletion-trxs-P1 plus deletion-trxs-P2 and deletion-trxs-P3 plus deletion-trxs-P4 were used . The resulting products were a 446-bp fragment containing the upstream region of trx1 and a 435-bp fragment containing the downstream region of trx1 . The resulting products were a 451-bp fragment containing the upstream region of trx2 and a 437-bp fragment containing the downstream region of trx2 . The resulting products were a 536-bp fragment containing the upstream region of trxlp and a 497-bp fragment containing the downstream region of trxlp . The fragments were cloned into the sacB suicide vector pDMK , linearized with Bgl II and Sph I , and the correct plasmids were introduced into E . coli CC118 λpir . Single-crossover mutants were obtained by conjugal transfer of the resulting plasmid into wild-type E . piscicida ( EIB202 ) . Deletion mutants were screened on 10% sucrose-tryptic soy agar ( TSA ) plates . All the mutants were confirmed by PCR amplification of the respective DNA loci , and subsequent DNA sequencing of each PCR product . ASK1-TBD and Trxlp were ligated into pET28a using the Nco I and Xho I sites , and the ASK1-TBD contains a sequence of 6 × His-tag at the C-terminal site , while the Trxlp contains a myelin basic protein ( MBP ) tag . The site-directed mutagenesis of ASK1-TBD or Trxlp was introduced using normal sequencing primers adjacent to multiple clone sites of plasmid as flanking primers . The recombinant plasmids containing genes were transformed into the E . coli strain BL21 ( DE3 ) , and the expressing proteins were water soluble after the induction of IPTG . All primers used for the construction of mutants are listed in S5 Table . The DNA fragment containing the ETAE_2186 gene , amplified by PCR from the vector pET28a-his-ETAE_2186 ( forward primer: 5’-GCGCGGATCCGTAGAGCCG GCCCTATAGCGACG-3’ , reverse primer: 5’-GCGCCTCGAGTTAGCGGGTCAGA AAGTCAG-3’ ) was inserted into the pET28b-His Sumo vector via the restriction sites BamH1 and Xho1 . The ligated plasmid was then transformed into E . coli BL21 ( DE3 ) cells . The resulting strain was grown to mid-log phase and then induced with 0 . 1 mM IPTG at 16°C for 16 h . Cells were collected by centrifugation and the pellet was resuspended in lysis buffer ( 50 mM Tris-HCl , pH8 . 0 , 400 mM NaCl , 10% glycerol , 2 mM 2-mercaptoethanol and protease inhibitor ) . Resuspended cells were lysed by sonication and cleared by high speed centrifugation at 40 , 000 g for 1 h . The protein was purified using Ni-NTA agarose beads ( GE ) , followed by enzyme digestion with Ulp1 to remove the His-Sumo tag . The digested product was further purified by gel-filtration chromatography ( Hiload Superdex 75 ) and passed through Ni-NTA beads again to remove residual His-Sumo contamination . The buffer for gel-filtration chromatography buffer contains 25 mM Tris-HCl , pH8 . 0 , 150 mM NaCl . The purified protein was concentrated to 22 mg/ml and store at -80°C . Crystal screenings were performed with Hampton screening kits by sitting-drop-vapor-diffusion at 20°C . Trxlp was crystallized in precipitant/well solutions: 30% PEG-8000 , 100 mM sodium acetate , pH 6 . 5 , 200 mM lithium sulfate . All crystals were gradually transferred into harvesting solutions ( precipitant solution and 25% glycerol ) before being flash-frozen in liquid nitrogen . Datasets were collected under cryogenic conditions ( 100 K ) at the Shanghai Synchrotron Radiation Facility ( SSRF ) beamlines BL19U1 , and were processed by HKL3000_ENREF_22 [45] . The crystal belongs to space group P6122 with cell dimension a = b = 34 . 263 Å , c = 264 . 649 Å . There is one molecule in an asymmetrical unit . The structure was solved by MR_Rosetta in Phenix package using Thermus thermophilus thioredoxin ( PDB:2YZU ) as the initial model . Iterative cycles of refinement and modeling were carried out using Phenix [45] and Coot [46] . All the crystal structural figures were generated using PyMOL [47] . The coordinates have been deposited in the RCSB PDB under the code PDB: 5ZF2 . Acquiring sequences from NCBI databanks and building alignment by manual adjustments based on structural alignments generated by the DNAman server . A CRISPR/Cas9 gRNA expression vector ( pSpCas9 ( BB ) -2A-Puro , #48138 ) was obtained from Addgene . The ASK1-KO target sequences were 5′-CACCGGCCGGG CAGCTTCTGGAACG-3′ , and 5′-AAACCGTTCCAGAAGCTGCCCGGCC-3′ . To generate ASK1-KO HeLa cell lines , DNA-In CRISPR Transfection Reagent ( MTI-GlobalStem ) was used to transfect the plasmids . Two to three days later , anti-puromycin cells were screened and cultured in DMEM complete medium containing 2 μg/ml puromycin . The puromycin resistant cells were series diluted and seeded into the 96-well plate . The ASK1-KO single clones were identified by immunoblotting with anti-ASK1 antibody . J774A . 1 cell was infected as described previously with slight adjustments [24] . Briefly , E . piscicida EIB202 was grown overnight in tryptic soy broth ( TSB ) at 30°C with shaking and then diluted into fresh TSB with shaking at 30°C until the optical density at 600 nm reached 0 . 8 . Harvested bacteria in phosphate-buffered saline ( PBS ) suspensions were added to macrophage cells at a multiplicity of infection ( MOI ) of 10:1 in DMEM containing 10% ( vol/vol ) FBS ( growth medium , GM ) . Plates were then centrifuged at 600 g for 10 min , and gentamicin ( 100 μg/ml ) was added 2 h after infection for 30 min to kill extracellular bacteria , followed by GM containing 10 μg/ml gentamicin for the remainder of the experiment at 30°C incubator . At indicated time points , the supernatant containing released bacteria was harvested . The supernatant was centrifuged at 600 g for 5 min to discard the cellular debris , and the harvested supernatant was further centrifuged at 13 , 000 g for 10 min to collect the macrophage-released bacteria . DMEM-cultured E . piscicida EIB202 was prepared as negative control . The macrophage-released and DMEM-cultured bacteria were prepared as described above [21] . RNA of both samples was extracted by using an RNA isolation kit ( Tiangen , Beijing , China ) . One microgram of each RNA sample was used for cDNA synthesis with the FastKing One Step RT-PCR Kit ( Tiangen ) and quantitative real-time PCR ( RT-qPCR ) was performed on an FTC-200 detector ( Funglyn Biotech , Shanghai , China ) by using the SuperReal PreMix Plus ( SYBR Green ) ( Tiangen ) . The gene expression of bacterial Trxlp was performed for three biological replicates , and the data for each sample were expressed relative to the expression level of the 16S gene by using the 2-ΔΔCT method . The gene expression of TNF-α , IL-10 , IL-6 , IL-8 and IFN-γ genes was performed for three biological replicates , and the data for each sample were expressed relative to the expression level of the β-actin gene by using the 2-ΔΔCT method . For Trx family protein secretion analysis , wild type and Trx-HA family protein-expressing E . piscicida EIB202 strains were grown overnight in TSB medium and subcultured 1:100 in fresh DMEM and grow for an additional 15 h . L-Arabinose was added to induce the expression of Trx-1 , Trx-2 and Trxlp-HA when OD600 was 0 . 6 . To ensure that protein from equal numbers of cells was analyzed , protein samples were adjusted to a volume in which 1 ml of culture corresponds to OD600 = 1 . Bacteria were collected in 50 ml tubes , and centrifuged at 5 , 000 g for 10 min at 4°C . Extracellular proteins were obtained by ultrafiltration from supernatants , which were filtered through a 0 . 22 μm filter membrane unit ( Millipore , Darmstadt , Germany ) with a 10 kDa molecular weight cut-off Amicon Ultra-15 centrifugal filter device ( Millipore ) . 150 micrograms of protein were boiled for 10 min in SDS sample buffer before each protein mixture was subjected to SDS-PAGE or stored at -20°C before Western blot analysis . To explore the secretory manner of protein , wild-type , ΔT3SS and ΔT6SS E . piscicida overexpressing with Trxlp-HA plasmid were cultured as indicated above . The protein samples were prepared and determined in the same method described above . OMVs were fractionated by density gradient ultracentrifugation with OptiPrep ( Sigma ) as described [26] . Briefly , EIB202 containing Trxlp-HA plasmid were grown overnight in TSB medium and then subcultured 1:100 in fresh DMEM and grown for an additional 12 h . Bacteria were removed by centrifugation ( 5 , 000 g , 10 min , 4°C ) and the DMEM supernatants were filtered through a 0 . 45 μm filter membrane unit ( Millipore ) with a 10 kDa molecular weight cut-off Amicon Ultra-15 centrifugal filter device . OMVs were collected from the filtered supernatants by ultracentrifugation ( 284 , 000 g , 1 . 5 h , 4°C ) in a CP-RX80 ( Hitachi , Japan ) and the OMV pellets were resuspended in 100 μl PBS , OMV-free supernatants were obtained from OMV supernatants . The protein samples were prepared and determined in the same method described above . The amounts of OMV were identified by the anti-OmpA antibody , and anti-RNAP antibody ( RNA polymerase ) was used to detect the bacteria . Infected HeLa cells were fractionated as reported previously with minor modifications [24] . Briefly , HeLa cells were seeded on 6-well/24-well culture dish at ( 35–40 ) ×104/ ( 7 . 5–10 ) ×104 per well before infection . Overnight cultured E . piscicida strains were diluted 1:100 into fresh DMEM with Amp and Col antibiotics , grown for 9–12 h , and then added into HeLa cells at a MOI of 100 in DMEM with arabinose , and incubated for 3 h . The culture dishes were then washed three times with pre-warmed PBS , and the medium was replaced with pre-warmed DMEM with 5% FBS ( with arabinose ) and 10 μg/ml gentamicin for another 5 h . The supernatant and cytosolic fraction from infected HeLa cells was harvest and analyze as described [24] . Wild-type E . piscicida were electroporated with pUTt0456-Trxs-HA plasmids to express Trxs-HA protein constitutively . HeLa cells were seeded onto 24-well plates containing sterile coverslips and cultured overnight . Following infection with above strains and incubation with gentamicin , the cells were washed with PBS and then fixed in 4% ( wt/vol ) paraformaldehyde for 10 min at room temperature . Fixed cells were washed in PBS and permeabilized with 0 . 1% Triton X-100 for 5 min at room temperature . After being washed with PBS , blocking of non-specific binding was achieved by placing the coverslips in the fetal bovine serum ( FBS ) for 20 min at room temperature . After blocking , a dilute solution of anti-HA antibody ( Molecular Probes ) was incubated with the coverslips under 4°C overnight . After rinsing the membrane to remove unbounded primary anti-HA antibody , the coverslips were exposed to secondary antibody for one hour and the nuclei were stained with 4 , 6-diamidino-2-phenylindole ( DAPI; Sigma ) for 10 min at room temperature . Fixed samples were viewed on a Nikon A1R confocal microscope , and the images were analyzed using ImageJ ( NIH ) . The translocation of translational protein fusions between TEM1 and Trxlp were evaluated by the detection of β-lactamase activity in infected HeLa cells as previously described [22] . Briefly , TEM1 fusions ( pCX340-trx1 , -trx2 , or -trxlp ) were introduced into wild-type , ΔT3SS or ΔT6SS E . piscicida by electroporation . Bacteria were grown in TSB overnight at 30°C , then diluted into DMEM and grown standing at 30°C until OD600 reached 0 . 8 . HeLa cells were then infected with strains harbouring the TEM1 fusions at a MOI of 100 . Infected cells were centrifuged at 400 g for 10 min to initiate bacterial-cell contact followed by incubation at 35°C for 3 h after which the cells were washed 3 times and incubated with fresh DMEM without serum for another 4 h . At this time point , cells were washed three times with DMEM and loaded with the fluorescent substrate CCF2/AM ( LiveBLAzer-FRET B/G loading kit; Invitrogen ) in the β-lactamase loading solution supplemented with 15 mM Probenecid ( Invitrogen ) . Cells were incubated in dark for 120 min at room temperature and then observed under a Nikon A1R confocal microscope . ASK1-TBD ( sequence 88–302 ) fragment was amplified from mammalian cells and identified by sequencing . HA-tagged ASK1-TBD expression plasmids in pCDH vector were constructed by introducing an HA epitope sequence at the C-terminal of ASK1-TBD by homologous recombination PCR ( Vazyme , Product code C112 ) . A GFP tag was inserted at the C-terminus of Trxlp in pCDH by PCR . HA-tagged ASK1-TBD ( 4 μg ) and GFP-Trxlp ( 6 μg ) expressing vectors were cotransfected into HEK293T cells in 10 cm dish with 10 ml medium by calcium phosphate transfection standard procedure . For immunoprecipitation , cells were lysed on ice using cell lysis buffer containing 20 mM Tris , 100 mM KCl , 0 . 1% NP-40 , 1 mM EDTA , 10% glycerol , 10 mM tetrasodium pyrophosphate , fresh cocktail . The lysates were divided and incubated with HA beads , and 6 h later , HA beads were washed with cell lysis and wash buffer ( 20 mM Tris , 150 mM KCl , 0 . 5% NP-40 , 1 mM EDTA , 1 ml EDTA , 10% glycerol , 10 mM tetrasodium pyrophosphate and fresh cocktail ) each for three times . Then , the beads were immunoblotted with either anti-GFP or anti-HA antibody . The proteins were detected with the ECL system . ASK1-N terminal ( sequence 1–400 ) fragment was amplified from mammalian cells and identified by sequencing . HA- and Flag-tagged ASK1-N terminal expression plasmids in pCDH vector were constructed by introducing the tag sequence at the C-terminal of ASK1-N terminal by homologous recombination PCR ( Vazyme , Product code C112 ) . The ASK1-N-HA , ASK1-N-Flag and Trxlp-HA were transfected into HEK293T cells as above . Thirty-six hours post transfection , the homogenate was prepared for the immunoprecipitation assay . The beads were immunoblotted with either anti-Flag or anti-HA antibody . The proteins were detected with the ECL system . DNA encoding N-terminal fragments of human ASK1 ( residues 88–302 ) were ligated into pET28a using the NcoI and XhoI sites , and 6 Histidine sequence was added by design of primers for proteins purification . Trxlp cloned from E . piscicida DNA was expressed with MBP tag ( in pET28a ) in E . coli . The recombinant tagged-fusion protein expression was induced by IPTG at 16°C for 16 h and purified from E . coli BL21 ( DE3 ) cells . rASK1-TBD-MBP protein was washed with MBP binding buffer ( 50 mM Tris-HCl ( pH 8 . 0 ) , 400 mM NaCl , 5 mM DTT , 10% ( w/v ) glycerol ) , eluted against MBP binding buffer containing 20 mM maltose and purified using Amicon Ultra centrifugal filters ( UFC501008 , Amicon Ultra-0 . 5 Centrifugal Filter Unit with Ultracel-10 membrane ) . Trxlp-His protein was gradiently eluted with His binding buffer ( Na2HPO4 ( 1 . 4 g/L ) , NaH2PO4 ( 1 . 216 g/L ) , NaCl ( 29 . 2 g/L ) , 5 mM DTT , imidazole ) , and Trxlp-his protein in 200 mM imidazole were concentrated using Amicon Ultra centrifugal filters ( UFC501008 , Amicon Ultra-0 . 5 Centrifugal Filter Unit with Ultracel-10 membrane ) by centrifugation ( ≤ 5 , 000 g , 4°C ) . All ASK1-TBD mutants and Trxlp mutants were generated by PCR and mutations were confirmed by sequencing . The indicated mutant proteins were expressed and purified as described above . The free thiol groups of purified Trx1 , Trx2 and Trxlp of E . piscicida were analyzed by Ellman’s detection with DTNB according to previous study [28] . The indicated thioredoxin’s catalytic reduction experiment was performed according to previous studies with slightly changes [29–30] . Briefly , the increase in turbidity at 650 nm is plotted against the reaction time . The assay mixtures contained 170 uM insulin , 2 mM DTT in 50 mM Tris-HCl ( pH 7 . 4 ) , 1 mM EDTA ( pH 7 . 0 ) , and the same concentrations of Trx protein , or ddH2O as negative control . Purified ASK1-His and Trxlp-MBP proteins were determined with Bradford standard method . Purified MBP fusion proteins immobilized on the MBP beads were incubated on ice for 30 min , and washed with MBP binding buffer for three times . Then ASK1-TBD-His protein was added into prepared Trxlp-MBP beads , incubated for another 4 h at 4°C , washed with MBP binding buffer to get rid of protein impurity , and finally eluted with MBP binding buffer containing 20 mM maltose . The eluent was detected with SDS-PAGE . Similarly , purified Trxlp His fusion proteins immobilized on His beads were incubated on ice for 30 min , washed with His binding buffer for three times . then Trxlp-MBP protein was added into prepared rASK1-TBD-His beads , incubated for another 4 h at 4°C , washed with His binding buffer containing 100 mM imidazole to remove the unbinding proteins , and finally eluted with His binding buffer containing 200 mM imidazole . The eluent was detected with SDS-PAGE . The binding complexes were determined by SDS-PAGE , and the lower part of the SDS-PAGE was cut out and probed with anti-His antibody or anti-MBP antibody . The protein was detected with the ECL system . HeLa cells were seeded at a density of 5 × 106 cells per well in 12-well plates and cultured overnight . Before infection , the culture medium was changed to serum-free DMEM for 12–16 h at 30°C . ZF4 cells were seeded at a density of 5×106 cells per well in 12-well plates and cultured overnight . Before infection , the culture medium was changed to serum-free DF12 for 12–16 h at 28°C . Wild-type E . piscicida and indicated mutant strains were cultured as described above , and infected at an indicated MOI . At indicated time , the infected cells were detected by immunoblotting or RT-PCR . HEK293T cells were transfected using standard calcium phosphate method [48] , and the medium were changed to serum-free DMEM for 12–16 h before TNF-α stimulation . Cells were lysed in 50 mM Tris , 150 mM NaCl , 1% Triton X-100 , and 1 mM EDTA with pH 7 . 4 . Lysates were mixed with protein loading buffer , boiled , and centrifuged . 10 μl of the cell lysate was separated by SDS-PAGE on a 12% gel and transferred to PVDF membrane ( Millipore ) , then probed with specific antibodies , and antibody binding detected by chemiluminescence . Antibodies for HA ( 0906–1 ) , Flag ( M1403-2 ) and β-actin ( M1210-2 ) were purchased from HUABIO at 1:5000 . Antibody for RNAP ( sc-101597 ) was purchased from Santa Cruz at 1:1000 . Tubulin ( AF1216 ) and calnexin ( AC018 ) antibodies were obtained from Beyotime Biotechnology at 1:1000 . Anti-His antibody ( AB102‐02; TianGen Biotech ) or anti-MBP antibody ( A00190‐100; Genscript Technology ) was used as primary antibodies for immunoblotting . Antibodies for IκBα ( 4814 ) , Erk1/2 ( 9107 ) , phospho-Erk1/2 ( 9101s ) , p38α ( 9218 ) , phospho-p38α ( 9215s ) , JNK ( 9252 ) , phospho-JNK ( 9251S ) , phospho-MEK1/2 ( 9121 ) , phospho-MKK3/6 ( 9231S ) , phospho-MKK4 ( 9156S ) , phospho-MKK7 ( 4171S ) and phospho-ASK1 ( 3765 ) were purchased from Cell Signaling Technology at 1:1000 . ASK1 antibody ( ET1608-54 ) was from HUABIO at 1:1000 . HRP-conjugated goat anti-mouse ( A0216 ) and anti-rabbit IgG ( A0208 ) antibodies were purchased from Beyotime Biotechnology at 1:5000 . An MO ( Gene Tools ) was designed to target a site in ask1 to block its translation ( 5’- TGAAGCGACACTCAGCAGTAGCTG-3’ ) , and a corresponding 5-base mismatch oligonucleotide was used as a specificity control ( 5’-TcAAcCGACAgTCAGCAcTAGgTG-3’ ) . Embryos were injected as described above . MOs were used at 0 . 25 , 0 . 5 , and 1 ng per embryo , and the results of experiments using three batches of embryos were analyzed independently . Knockdown of ask1 was verified by immunoblotting with anti-ASK1 antibody . Zebrafish embryos were collected from a laboratory-breeding colony kept at 28°C on a 12:12 h light/dark rhythm as previously described [37] . Embryos were raised in Petri dishes with E3 medium ( 5 mM NaCl , 0 . 17 mM KCl , 0 . 33 mM CaCl2 , 0 . 33 mM MgSO4 ) containing 0 . 3 μg/ml methylene blue at 28°C . Zebrafish larvae were maintained until 3-day post fertilization ( dpf ) , at which time all were euthanized by 4 g/L buffered tricaine ( MS-222 , ethyl 3-aminobenzoate methanesulfonate , Sigma-Aldrich ) in accordance with ethical procedures . For microinjection experiments , the bacteria were prepared as described in culture conditions . Injections were done using pulled borosilicate glass microcapillary injection needles ( Sutter ) and a Milli-Pulse Pressure Injector ( ASI ) . Prior to injections , embryos of 3 dpf were manually dechorionated and anesthetized with 200 mg/L buffered tricaine ( MS-222 ) . Afterwards embryos were aligned on an agar plate and injected with 1 nl of the indicated E . piscicida suspension into the yolk sac . Prior determination of the injected volume was performed by injection of a droplet into mineral oil and measurement of its approx . diameter over a scale bar . After injections , infected larvae were allowed to recover in a Petri dish with fresh E3 medium for 15 min . Subsequently , larvae were transferred in 10 cm dish in groups of about 50 larvae in 15 ml E3 medium per dish , incubated at 28°C . The mortality rate and bacterial colonization at different time points were observed . The RNA of infected zebrafish larvae was extracted , and the expression of TNF-α , IL-10 , IL-6 , IL-8 and IFN-γ were detected by RT-PCR as described above . Statistical analysis was performed using GraphPad Prism program ( GraphPad Software ) . All data were representative of at least three independent experiments and were presented as mean ± SD ( standard deviation ) . Western bolts were analyzed by Quantity One Software , and Gauss model trace of each bands was calculated based on the equation Band-Gauss Model Bands . Differences between two groups were evaluated using Student’s t test . One-way ANOVA test was used to analyze differences among multiple groups . Differences in larvae survival were assessed using the log-rank ( Mantel-Cox ) test . Statistical significance was defined as * p<0 . 05 , ** p<0 . 01 , *** p<0 . 001 . | Thioredoxin ( Trx ) is universally conserved thiol-oxidoreductase that regulates numerous cellular pathways under thiol-based redox control in both prokaryotic and eukaryotic organisms . Despite its central importance , the mechanism of bacterial Trx recognizes its target proteins in host cellular signaling remains unknown . Here , we uncover a bacterial thioredoxin-like protein that can be translocated into host cells and mimic the endogenous TRX1 to target ASK1-MAPK signaling , finally facilitating bacterial pathogenesis . This work expands our understanding of bacterial thioredoxins in manipulating host innate immunity . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"fish",
"viral",
"transmission",
"and",
"infection",
"redox",
"signaling",
"hela",
"cells",
"molecular",
"probe",
"techniques",
"biological",
"cultures",
"immunoblotting",
"vertebrates",
"microbiology",
"animals",
"animal",
"models",
"osteichthyes",
"de... | 2019 | The Edwardsiella piscicida thioredoxin-like protein inhibits ASK1-MAPKs signaling cascades to promote pathogenesis during infection |
The loop-mediated isothermal amplification ( LAMP ) assay , with its advantages of simplicity , rapidity and cost effectiveness , has evolved as one of the most sensitive and specific methods for the detection of a broad range of pathogenic microorganisms including African trypanosomes . While many LAMP-based assays are sufficiently sensitive to detect DNA well below the amount present in a single parasite , the detection limit of the assay is restricted by the number of parasites present in the volume of sample assayed; i . e . 1 per µL or 103 per mL . We hypothesized that clinical sensitivities that mimic analytical limits based on parasite DNA could be approached or even obtained by simply adding detergent to the samples prior to LAMP assay . For proof of principle we used two different LAMP assays capable of detecting 0 . 1 fg genomic DNA ( 0 . 001 parasite ) . The assay was tested on dilution series of intact bloodstream form Trypanosoma brucei rhodesiense in human cerebrospinal fluid ( CSF ) or blood with or without the addition of the detergent Triton X-100 and 60 min incubation at ambient temperature . With human CSF and in the absence of detergent , the LAMP detection limit for live intact parasites using 1 µL of CSF as the source of template was at best 103 parasites/mL . Remarkably , detergent enhanced LAMP assay reaches sensitivity about 100 to 1000-fold lower; i . e . 10 to 1 parasite/mL . Similar detergent-mediated increases in LAMP assay analytical sensitivity were also found using DNA extracted from filter paper cards containing blood pretreated with detergent before card spotting or blood samples spotted on detergent pretreated cards . This simple procedure for the enhanced detection of live African trypanosomes in biological fluids by LAMP paves the way for the adaptation of LAMP for the economical and sensitive diagnosis of other protozoan parasites and microorganisms that cause diseases that plague the developing world .
Tsetse fly-transmitted African trypanosomes are major pathogens of humans and livestock . Two subspecies of Trypanosoma brucei ( T . b . rhodesiense and T . b . gambiense ) cause human African trypanosomiasis ( HAT , commonly called sleeping sickness ) . After replicating at the tsetse fly bite site , trypanosomes enter the hemolymphatic system ( early stage or stage 1 ) ( 5 , 9 ) . Without treatment , the parasites go on to invade the central nervous system ( CNS; late stage or stage 2 ) , a process that takes months to years with T . b . gambiense ( West and Central African HAT ) or weeks to months with T . b . rhodesiense ( East African HAT ) . The parasites cause a meningoencephalitis leading to progressive neurologic involvement with concomitant psychiatric disorders , fragmentation of the circadian sleep-wake cycle and ultimately to death if untreated ( 4 , 5 , 9 ) . A key issue in the treatment of HAT is to distinguish stage 1 from stage 2 disease , as the drugs used for the treatment of stage 2 need to cross the blood-brain barrier [1] , [2] . The most widely used drug is melarsoprol ( developed in 1949 ) , which is effective for T . b . gambiense and T . b . rhodesiense HAT , but unfortunately , melarsoprol leads to severe and fatal encephalitis in about 5–10% of recipients despite treatment for this condition [3] , [4] , [5] . Therefore , where HAT is endemic , accurate staging is critical , because failure to treat CNS involvement leads to death , yet inappropriate CNS treatment exposes an early-stage patient unnecessarily to highly toxic and life-threatening drugs . The diagnosis of HAT in the rural clinical setting , where most patients are found , still relies largely on the detection of parasitemia by blood smear and/or CSF microscopy [6] , [7] . While T . b . rhodesiense detection in blood is frequently successful , T . b . gambiense infections , which constitute over 90% of all HAT cases , typically show very low parasitemias , and concentration techniques such as centrifugation or mini-anion exchange columns are usually necessary [6] , [7] , [8] . Stage determination still relies on lumbar puncture to examine CSF for trypanosomes or white cell count/protein concentration suggestive of chronic meningoencephalitis . Threshold values for these parameters are controversial , with the conventional value for stage 2 ( >5 cells/µL ) now increased to >10 or even 20 cells/µL [9] . In summary , diagnosis and staging of HAT is currently time consuming , intensive and difficult . DNA-based diagnostic methods such as PCR and LAMP now offer greater sensitivity than existing diagnostic methods , detecting DNA from the equivalent of 0 . 01 parasites or less . Based on PCR protocols for HAT [10] , we described LAMP targeting the conserved paraflagellar rod A ( PRFA ) gene in all T . brucei subspecies and T . evansi [11] . LAMP is an isothermal DNA amplification method with excellent analytical sensitivity and specificity when employed for the detection of a variety of microorganisms ( reviewed in [12] ) , including human and animal infective African trypanosomes [11] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] . LAMP relies on autocycling strand displacement coupled to DNA synthesis by Bst DNA polymerase , a reaction similar to rolling-circle amplification [22] but with the added advantage that a heat-denaturing step is not necessary to initiate rounds of amplification . Specificity is dictated by four primers ( F3 , B3 , FIP and BIP ) , and the addition of two loop primers ( LoopF and LoopB ) significantly reduces the reaction time [23] . LAMP is cost-effective ( <1 US dollar/test ) , simple ( the isothermal reaction requires a simple heating device ) , and rapid ( within 60 minutes ) [12] , [24] . Furthermore , Bst DNA polymerase can be stored for weeks at ambient temperatures , a critical property where maintaining a cold chain is difficult [13] . Positive reactions are indicated by turbidity [25] , color changes after addition of hydroxynaphthol blue ( HNB ) [26] , or changes in fluorescence using indicator dyes [26] , [27] , [28] . Despite its advantages , the usefulness of LAMP for HAT diagnosis is handicapped in the clinical setting by its inability to directly detect live trypanosomes in blood or CSF below 1 parasite/µL ( 103 parasites/mL ) , the practical detection limit based on a 1 µL sample volume often used in LAMP or PCR assays . While sensitivity can be increased 5–10 fold by adding more sample volume , significant improvement in the assay system for the detection of live parasites in biologically relevant samples would clearly be of benefit for diagnosis . Here , we introduce a very simple modification to the LAMP assay recognizing multi-copy gene targets that can increase the analytical sensitivity for the detection of live parasites 100-fold or more .
Two LAMP primer sets were tested . Data on analytical sensitivity and specificity of a LAMP primer set for trypanosome DNA based on the multicopy ( approximately 500 copies ) repetitive insertion mobile element ( RIME ) of subgenus Trypanozoon ( GenBank Accession No . K01801 ) is well-documented [16] , [26] , [29] . Using between 2–4 µL sample , this pan-T . brucei assay is reported to detect DNA from ∼0 . 001 trypanosome [16] . LAMP primers based on the serum resistance associated ( SRA ) gene ( GenBank AJ560644 ) ( see Table S1 for gene sequence ) were designed using PrimerExplorer version 4 software ( http://primerexplorer . jp/e/ ) to create the basic F3 , B3 , FIP , BIP [30] and loop LF , LB [23] primers ( Fig . 1A ) . As this assay recognized more than the SRA gene , this primer set is designated PSEUDO-SRA . All RIME and PSEUDO-SRA LAMP primers were synthesized and HPLC purified . For comparison , we also used the SRA gene ( GenBank accession number Z37159 ) -specific LAMP assay [17] . Genomic DNA was prepared by using Qiagen DNeasy Blood & Tissue Kits . The LAMP reaction was performed as previously described [11] , [14] , [15] . Briefly , the reaction contained 12 . 5 µL of 2x LAMP buffer ( 40 mM Tris-HCl [pH 8 . 8] , 20 mM KCl , 16 mM MgSO4 , 20 mM [NH4]2SO4 , 0 . 2% Tween 20 , 1 . 6 M Betaine , 2 . 8 mM of each deoxyribonucleotide triphosphate ) , 1 . 0 µL primer mix ( 5 pmol each of F3 and B3 , 40 pmol each of FIP and BIP ) or 1 . 3 µL when LF and LB ( 20 pmol each ) were included , 1 µL ( 8 units ) Bst DNA polymerase ( New England Biolabs , Tokyo , Japan ) , 1 µL of template DNA . Final volumes were adjusted to 25 µL with distilled water . All reactions were conducted in 2 to 4 replicates and were monitored in real-time in a Loopamp® real-time turbidimeter LA320C ( Teramecs , Tokyo , Japan ) . The optimal reaction temperatures were 62°C ( RIME LAMP ) and 63°C ( PSEUDO-SRA LAMP ) . The reaction was terminated by increasing the temperature to 80°C for 5 min . In addition to turbidity , the amplified products were analyzed on 2% agarose gels using the E-Gel EX system with ethidium bromide or SYBR green incorporated into the gels ( Invitrogen ) , and/or after addition of hydroxynaphthol blue ( HNB ) [26] to enable visual detection . The HNB color change from violet to sky blue has been consistently interpreted by independent observers as the easiest to see [26] . Human CSF was obtained as discarded samples from The Johns Hopkins Hospital Microbiology laboratory with approval of the Johns Hopkins Medicine IRB . CSF were adjusted to contain either 1/20 volume deionized water ( untreated CSF ) or 1/20 volume 10% ( w/v ) Triton X-100 ( final concentration 0 . 5% Triton ) . A 10% ( w/v ) Triton X-100 stock solution was made by adding 1 g Triton X-100 to a final volume of 10 mL DNase/RNase free water ( Qiagen ) . We used bloodstream form T . b . rhodesiense IL1852 , a CSF isolate from a patient in Kenya [31] , [32] . Originally thought to be T . b . gambiense it has been reclassified as T . b . rhodesiense [33] based on the presence of the SRA gene [34] and the absence of the TgsGP gene [35] , [36] ( Fig . S1 ) . Human CSF was spiked with bloodstream form T . b . rhodesiense IL1852 and the samples serially diluted 1∶10 in CSF with or without 0 . 5% detergent to cover a range of parasite concentrations from 104 to 10−1 parasites/mL . After 60 min incubation at ambient temperature to allow for lysis , the LAMP assays were done using 1 µL CSF . In the field , biological samples are often shipped to another geographical site for later analyses . They are often preserved by spotting on paper cards designed for short-term protein , RNA and DNA storage ( 2 weeks at ambient temperature ) such as Whatman Protein Saver 903 , or long-term ( years ) DNA storage/archiving on Whatman FTA cards . To simulate these conditions , human blood obtained as discarded samples from The Johns Hopkins Hospital Microbiology laboratory with approval of the Johns Hopkins Medicine IRB was spiked with T . b . rhodesiense ( 104 to 10−2 parasites/mL ) . Protein Saver 903 cards were pretreated with 50 µL 0 . 5% Triton X-100 , which was sufficient to fill the designated circle on the cards , and dried overnight prior to whole blood spotting . For assay standardization , three 2 mm punches made from the 1 cm2 dried blood spots ( DBS ) and DNA was extracted using standard methods [37] . The LAMP assays were done using 1 µL DBS DNA template . Alternatively , untreated and 0 . 5% Triton X-100 treated trypanosome-spiked blood were spotted on untreated Protein Saver 903 cards and dried overnight with subsequent DBS DNA extraction as above .
Based on experiments repeated at least 3 times , LAMP assays successfully amplified T . b . rhodesiense DNA within 55–60 min at 62°C ( RIME LAMP ) or 63°C ( PSEUDO-SRA LAMP ) . As reported previously [16] , we found that RIME LAMP detected 0 . 1 fg genomic DNA ( 0 . 001 parasite ) from T . b . rhodesiense IL1852 ( not shown ) . PSEUDO-SRA LAMP was as sensitive and reliably detected 0 . 1 fg ( 0 . 001 parasite ) or less T . b . rhodesiense IL1852 genomic DNA ( Fig . 1B and 1C ) . Nonetheless , when using the SRA gene specific LAMP assay [17] the detection limit for T . b . rhodesiense IL1852 genomic DNA was 0 . 1–1 . 0 pg ( 1–10 parasites ) , comparing favorably to reported values [17] . The standard curves with PSEUDO-SRA LAMP seem to display biphasic kinetics with an early initial phase ( 15–20 min ) followed by a late second phase ( 35 and 55 min ) with a break point around 1 . 7 fg DNA ( Fig . 1B ) suggesting that it targets other genomic components besides the SRA gene . As SRA is a truncated VSG , it is likely that the PSEUDO-SRA LAMP is amplifying other VSG sequences , albeit not efficiently ( see below ) . Although the PSEUDO-SRA LAMP primer sequences were verified as being unique by BLAST analysis of the T . b . brucei TREU 927 genome sequence and the VSG database ( TriTrypDB: http://tritrypdb . org/tritrypdb/ ) , VSG repertoires are diverse between strains , and we were unable to assess the primers against sequences of the full IL1852 VSG repertoire as its genome has not been sequenced . The PSEUDO-SRA LAMP was specific and recognized DNA equally well from other T . b . rhodesiense strains ( LouTat 1A , GYBO , IL1501 ) , but did not recognize DNA from T . b . gambiense isolates ( IL 3258 , DAL 972 , DAL 072 , DAL 069 , IPR SG-1020 , FONT l′93 , JUA , MOS , MA 002 ) ( not shown ) . It also recognized T . b . brucei strain 927 genomic DNA at very high concentrations ( i . e . >1 ng DNA/µL equivalent to >104 parasites/µL ) , but it was specific for T . b . rhodesiense at the concentrations tested ( 10 pg to 0 . 1 fg DNA/µL equivalent to 102 to 10−3 parasites/µL ) . Negative controls included the eukaryotic protozoan parasites Babesia microti , Plasmodium falciparum , Plasmodium ovale , and Toxoplasma gondii , as well as DNA from clinical samples or spiked blood samples , such as Borrelia burgdorferi , Borrelia crocidurae , Enterococcus spp . , Ehrlichia chaffeensis , Escherichia coli , Pseudomonas aeruginosa , Rickettsia parkeri , Staphylococcus spp . , and DNA from mouse and human blood . Furthermore , using PSEUDO-SRA LAMP under carefully controlled conditions , no false positives were found when DNA from 192 normal human CSF samples was tested . Although it is possible to detect very low parasite numbers using Psuedo-SRA LAMP , the assay's sensitivity is a potential drawback because of risk for amplicon contamination . Therefore , until more validation is done , we do not propose PSEUDO-SRA LAMP for diagnosis of T . b . rhodesiense . However , the range of sensitivity made it an ideal choice to study the effects of detergent on increasing the ability of LAMP to detect live parasites in biological samples . To mimic a clinical situation , we first tested RIME LAMP and PSEUDO-SRA LAMP on human CSF spiked with live T . b . rhodesiense IL1852 and analyzed the reaction products on agarose gels , HNB reaction , and/or real-time LAMP based on turbidimetric readings . As predicted , the LAMP assays had a detection limit of 103 parasites per mL based on 1 µL assay samples for RIME ( Fig . 2 ) and PSEUDO-SRA LAMP ( Fig . 3 ) . While sensitivity could be increased up to 10 fold by increasing CSF sample volume to 10 µL ( not shown ) , the presence of detergent ( i . e . 0 . 5% Triton X-100 ) alone added to the CSF samples improved detection to 10 and 1 parasite/mL , representing a 100 to 1000-fold increase in RIME LAMP and PSEUDO-SRA LAMP assay analytical sensitivity , respectively ( Figs 2 and 3; Table 1 ) . Release of parasite DNA by 0 . 5% Triton required between 30 and 60 min incubation . The transport and storage of DBS or CSF on filter paper cards is a common practice in the field . DBS on Whatman Protein Saver 903 cards are used for parasite pathogen detection ( DNA , RNA and/or protein ) and genotyping [38] , [39] , [40] , [41] , [42] , [43] . Depending on the paper matrix , DNA , RNA and/or protein to be tested are first extracted from defined diameter punches ( e . g . 2 mm ) and 1–5 uL are assayed . Assay sensitivity for trypanosomes is limited by the stoichiometric presence of the parasite in the assayed sample . Analytical sensitivity is further reduced since sample volumes in filter paper punches represent <1% of the total captured on the paper itself [44] . We used parasite-spiked human blood spotted on dry Protein Saver 903 cards pretreated with detergent . Remarkably , the presence of detergent greatly enhanced LAMP detection limits for parasite DNA by about 100 fold for RIME and PSEUDO-SRA LAMP ( Figs 4 and 5; Table 1 ) . Enhanced detection sensitivity was also found when T . b . rhodesiense IL1852 DNA was extracted from DBS from Protein Saver 903 cards containing normal or detergent-treated parasite-spiked human blood ( Figs S2 and S3 ) . In general , replicates were more reproducible in assays where the detergent was present in the paper . The presence of detergent had no effect on analytical sensitivity by HNB [26] , confirming its use for easy , inexpensive , accurate , and reliable field detection of LAMP-amplified DNA . As with any DNA amplification method , standard precautions for avoiding template contamination [45] also apply for LAMP-based assays . It has been shown that sensitivity , including detection of type 1 T . b . gambiense [20] , can be greatly enhanced after heat denaturing the samples before LAMP assay [16] , [17] , [20] . However , this procedure is less convenient than simply incubating samples at ambient temperature with detergent or allowing samples to dry on detergent-pretreated filter cards . Aerosol effects by heating the samples could also increase the risk of cross-contamination prior to addition of the reaction mixture . Furthermore , the extra steps required for techniques such as quantitative buffy coat , microhematocrit centrifugation ( mHCT ) , mini-anion-exchange centrifugation technique ( mAECT ) used to concentrate the parasites from blood or CSF [6] , [7] also increase contamination risk . The addition of samples directly in the reaction helps reduce contamination . Recent findings by Deborggraeve et al . [46] suggest that while PCR performed better than , or similar to current parasite detection techniques for T . b . gambiense sleeping sickness diagnosis and staging , it cannot be used for post-treatment follow-up because of persistence of living or dead parasites or their DNA after successful treatment . The use of LAMP on serially diluted sample in the absence and presence of detergent could be useful for differentiating between these scenarios; large difference might indicate a recent infection with small differences indicating persistent or relapse infection . While we have not yet optimized conditions with regards to detergent concentration or class ( nonionic , ionic or zwitterionic ) , our preliminary evidence supports the concept that a detergent such as Triton X-100 can be used in a variety of ways to enhance the analytical sensitivity of multi-copy gene LAMP-based assays for the detection of intact African trypanosomes in blood and CSF approximately approaching or reaching the detection limits of LAMP for genomic DNA . In addition to LAMP , the implications of these findings are far reaching and should also be applicable for improved lateral-flow dipstick methods recently introduced [47] , PCR , or other nucleic acid amplification-based [Recombinase Polymerase Amplification ( TwistDX ) , Strand Displacement Amplification ( Probetec ET , Becton-Dickinson ) , Nucleic Acid Sequenced Base Amplification ( Primer Biosoft International ) ] technologies where microbial pathogen , including protozoan parasite ( e . g . Plasmodium ) DNA/RNA could be easily released by detergents . Unlocking the potential power of LAMP for accurate HAT diagnosis presents an excellent option for the administration of effective anti-trypanosome treatment . In summary , the procedure paves the way for the adaptation of LAMP and similar technologies as simple cost-effective diagnostics for intact African trypanosomes in humans , animals and tsetse flies , and also for other protozoan parasites and microorganisms that cause diseases that plague the developing world . | Human African trypanosomiasis or sleeping sickness is a fatal disease ( if untreated ) spread by bloodsucking tsetse flies . Trypanosome parasites first enter the blood and lymph and eventually invade the brain . In rural clinical settings , diagnosis still relies on the detection of these microbes in blood and cerebrospinal fluid ( CSF ) by microscopy . LAMP , or loop-mediated isothermal amplification of DNA , is a technique that can specifically detect very small amounts of DNA from an organism . It is similar to PCR , the polymerase chain reaction , another DNA amplification technique widely used for diagnosis of infectious diseases . LAMP's advantages are that the reaction works at one temperature , whereas PCR needs a thermocycler , and LAMP is not affected by blood components that inhibit PCR . We show that by simply adding detergent during sample preparation , the analytical sensitivity of LAMP targeting many gene copies is greatly improved , presumably because DNA is released from the pathogen cells and dispersed through the sample . To demonstrate proof of principle , we used pathogenic trypanosomes in different human body fluids ( CSF or blood ) , but this simple modification should be applicable for diagnosis of other microbial infections where cells are sensitive to detergent lysis . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"and",
"Discussion"
] | [
"biology"
] | 2011 | Using Detergent to Enhance Detection Sensitivity of African Trypanosomes in Human CSF and Blood by Loop-Mediated Isothermal Amplification (LAMP) |
The effects of genetic variants on phenotypic traits often depend on environmental and physiological conditions , but such gene–environment interactions are poorly understood . Recently developed approaches that treat transcript abundances of thousands of genes as quantitative traits offer the opportunity to broadly characterize the architecture of gene–environment interactions . We examined the genetic and molecular basis of variation in gene expression between two yeast strains ( BY and RM ) grown in two different conditions ( glucose and ethanol as carbon sources ) . We observed that most transcripts vary by strain and condition , with 2 , 996 , 3 , 448 , and 2 , 037 transcripts showing significant strain , condition , and strain–condition interaction effects , respectively . We expression profiled over 100 segregants derived from a cross between BY and RM in both growth conditions , and identified 1 , 555 linkages for 1 , 382 transcripts that show significant gene–environment interaction . At the locus level , local linkages , which usually correspond to polymorphisms in cis-regulatory elements , tend to be more stable across conditions , such that they are more likely to show the same effect or the same direction of effect across conditions . Distant linkages , which usually correspond to polymorphisms influencing trans-acting factors , are more condition-dependent , and often show effects in different directions in the two conditions . We characterized a locus that influences expression of many growth-related transcripts , and showed that the majority of the variation is explained by polymorphism in the gene IRA2 . The RM allele of IRA2 appears to inhibit Ras/PKA signaling more strongly than the BY allele , and has undergone a change in selective pressure . Our results provide a broad overview of the genetic architecture of gene–environment interactions , as well as a detailed molecular example , and lead to key insights into how the effects of different classes of regulatory variants are modulated by the environment . These observations will guide the design of studies aimed at understanding the genetic basis of complex traits .
We are rapidly approaching an age when genomic sequence will be used to inform life decisions . The extent to which lifestyle choices will change how our genetic blueprint is expressed , however , depends on the presence of gene–environment interactions: the phenomenon where the effect of a genetic variant differs in multiple environments . In humans , gene–environment interactions have been reported for many diseases , including heart disease ( reviewed in [1] ) , depression [2] , and cancer [3] . More often , however , studies either fail to see these effects or they are difficult to reproduce . Studies in model and agricultural organisms have been more promising , particularly in experimental crosses where gene–environment interaction can be studied more easily on a genome-wide level , as opposed to only targeting candidate genes . When quantitative trait linkage analysis has been performed for the same trait in multiple environments , different loci are often found in the different environments [4] , and when tested explicitly , these loci often display gene–environment interaction [5 , 6] . However , these studies are limited in their scope because they either examine one or a few phenotypes or restrict their analysis to loci that were initially found within a single condition . In humans , as well as in experimental systems , tracking down the molecular mechanisms of gene–environment interaction has proved difficult . Recently , gene transcript abundance has been used as a model to study the genetics of thousands of quantitative traits at a time [7 , 8] . The studies have targeted organisms such as yeast [9] , humans [10–13] , mice [14–16] , rat [17] , and Arabidopsis [18] . Because so many traits can be studied at once , these studies have been able to elucidate trends in difficult-to-study phenomena , such as gene–gene interaction [19 , 20] and genetic complexity [21] . Gene–environment interaction recently has been studied using transcript levels in yeast [22] and in worms [6] . Landry et al . ( 2006 ) compared six strains in three conditions and found 221 transcripts showing gene–environment interaction [22] , suggesting that there is much phenotypic diversity at the strain level in how yeast responds to its environment . Li at al ( 2006 ) mapped quantitative trait loci responsible for expression differences in two different temperatures in Caenorhabditis elegans [6] . They identified 197 linkages that showed gene–environment interaction , suggesting that studying gene–environment interaction of expression phenotypes using model organisms in controlled conditions should prove fruitful . We have used a large family of yeast segregants to characterize gene–environment interaction on a global scale . Two parental strains ( BY and RM , see Materials and Methods ) and 109 segregant strains from a cross between the parental strains were grown in two conditions—glucose and ethanol as carbon sources—and expression profiled using microarrays . When growing in glucose , yeast predominantly ferment glucose to ethanol . When the cells run low on glucose , they switch to a primarily respiratory state in which they metabolize ethanol [23] . The transcriptional state of the cells changes dramatically during this period [24 , 25] , with transcripts influencing growth and respiration particularly affected . This difference in the metabolic state may change how genetic variants influence traits , resulting in the observation of gene–environment interaction . By studying thousands of such traits , we sought to characterize the general genetic architecture of gene–environment interaction , and by using the molecular tools available in yeast , we were able to characterize some of the variants involved at the gene level .
We characterized the extent of gene–environment interaction influencing transcriptional phenotypes in the parental strains BY and RM by growing six independent cultures of each strain in glucose and in ethanol , and measuring gene expression with microarrays ( Figure 1 and Dataset S1 ) . We determined the influence of strain , condition , and the interaction between strain and condition ( here we use strain–condition interaction to distinguish effects due to overall strain differences in genetic background , as opposed to a specific gene or locus ) by using analysis of variance ( ANOVA ) . Of the 4 , 342 transcripts with high-quality data , 2 , 037 ( 47% ) , 2 , 996 ( 69% ) , and 3 , 448 ( 79% ) showed significant effects due to strain–condition interaction , strain , and condition , respectively ( see Figure 1B and 1C for strong examples and Table S3 for a full list ) . Transcripts were often influenced by multiple factors ( Figure S2A ) . We detected ten times as many transcripts showing strain–condition interaction as has been previously reported for gene expression in yeast [22] . This result is primarily due to the number of replicates we performed ( six ) ; an analysis with two replicates detected a number of interactions similar to the previous report . Strain–condition interaction effects accounted for 9% of the total variance explained over all transcripts , whereas strain and condition effects were larger , explaining 21% and 36% of the variance , respectively . The importance of each of these factors varied for individual transcripts , with each factor playing a dominant role for a subset of transcripts ( Figure 2 ) . Genetic correlations of transcript levels between the two conditions were on average low ( mean genetic correlation = 0 . 26 ) , with transcripts that show significant strain–condition interaction having lower genetic correlations than those that do not ( −0 . 07 vs . 0 . 56 , respectively; Mann-Whitney p < 10−15 ) . These genetic correlations are lower than those observed for outbred livestock populations [26 , 27] . Thus , as previously reported , strain [9] and condition [25] effects profoundly influence gene expression , but interaction effects between the two are also important . To understand the genetic basis of gene–environment interaction , we measured transcript expression in 109 segregants derived from a cross between BY and RM [21] . Each of the 109 previously genotyped segregants was expression profiled in both glucose and ethanol conditions on microarrays ( Dataset S2 ) . We performed linkage analysis on the transcript levels within each condition , and 3 , 997 and 3 , 489 linkages were observed in glucose and ethanol , respectively ( Figure S1 and Table S4 ) . We determined loci that show gene–environment interaction ( gxeQTL ) by performing linkage analysis on the difference between conditions ( for each segregant , the difference between conditions is the difference between expression in ethanol and expression in glucose; this type of measurement is also known as plasticity [28] and is similar to environmental sensitivity [29] ( see Text S1 ) ) . The difference between conditions was calculated for each segregant for 4 , 482 transcripts with high-quality data ( see Materials and Methods ) , and subjected to linkage analysis with 2 , 894 markers spaced throughout the genome . There were 1 , 382 transcripts that showed gene–environment interaction with at least one locus ( total of 1 , 555 linkages ) at a genome-wide 5% false discovery rate ( FDR ) , as determined by permutation ( see Figure 3 for an example ) . Transcripts that showed strain–condition interaction in the parents were more likely to show at least one linkage in the segregants ( relative risk [RR] = 1 . 8 , 95% confidence interval [CI] = 1 . 6–1 . 9 , χ2 = 154 , p = 3 × 10−35 ) , but there was not a one-to-one correspondence . For some transcripts , effects were revealed in the segregants that were not observed in the parents . For others , we were unable to map specific loci responsible for differences observed in the parents . Some of this may due to differences in power , but because of genetic complexity , we do not expect to map a locus for all transcripts that show significant genetic or gene–environment interaction effects in the parent strains . Previous work has also shown that effects that are revealed only in the segregants ( transgressive segregation ) are common in the genetic regulation of gene expression within a condition [21] . We think that these phenomena are likely to play a role here as well . Since transcripts derive from open reading frames ( ORFs ) with physical locations in the genome , we can make a distinction between linkages that are physically near the ORF ( local ) and those that are far away from the ORF ( distant ) . Local linkages are often due to variation at cis-acting sites [30–32] , whereas distant linkages likely result from variants that influence a trans intermediate , such as those resulting in a change in protein function or concentration . Most transcripts linked to distant loci . However , for gxeQTL , this was especially pronounced . gxeQTL were less likely to be local , with 172/1 , 555 ( 11% ) of all gxeQTL being local , as compared to linkages within either condition , where 22%–25% of all linkages were local ( Table 1 ) . This result agrees with previous reports that either directly addressed interaction or observed differences between conditions [6 , 10 , 15–17 , 33 , 34] , and suggests that changes in cis-regulatory sites are less condition dependent than those influencing trans-acting factors . As has been previously reported in mouse [14–16] and yeast [9] , distant linkages cluster in a small number of regions . We divided the genome into 10-cM–sized bins and counted the number of distant linkages observed in glucose , ethanol , or gxeQTL that fell within each of the bins ( Figure 4 ) . The majority of distant linkages fell into bins with a significant excess of linkages , with 81% ( 1 , 126/1 , 383 ) of distant gxeQTL occurring in 31/563 bins , or 5 . 5% of the genome . Significant bins that were located immediately next to each other were merged into a single peak to form 13 , 13 , and 15 peaks for glucose , ethanol , and gxeQTL , respectively ( Figure 4 and Table S1 ) . Some of these regions coincided with loci that have been characterized at the gene level , including AMN1 [35] , GPA1 [35] , MAT [19] , HAP1 [9] , and IRA2 ( this paper ) . Peaks identified in glucose did not coincide with those in ethanol ( Figure 4 ) . Thirteen peaks were found in each condition , but only seven were found in both . Two large peaks are found for gxeQTL , but are only observed within one condition , suggesting that the polymorphism likely acts in one condition , but not the other . We highlight these as striking examples of differences in genetic regulation across conditions . The first of these colocalizes with the gene encoding Gpa1 , the G-protein alpha subunit of the pheromone response signaling pathway [36 , 37] , which contains a polymorphism previously associated with transcripts involved in pheromone response [35] . In glucose , 49 transcripts show distant linkage here , whereas in ethanol , there are no distant linkages , suggesting that the polymorphism in GPA1 does not influence these transcripts in ethanol . When we asked where these transcripts linked in ethanol , we found that 12 showed linkage to eth12 , a peak that is also enriched for genes involved in the pheromone response . Given the overall rates of linkage to both loci , we would expect less than one overlap by chance , and the observed overlap of 12 genes is highly significant ( Fisher exact p-value = 8 × 10−19 ) . Within this region lies DIG1 , which codes for an inhibitor of the pheromone response/invasive growth transcription factor Ste12 [38 , 39] . Allele replacements of DIG1 in both backgrounds show that variants in DIG1 make a major contribution to the effects at this locus ( Figure S3 ) . This example highlights how environmental differences can change the relative importance of genetic variants in the same pathway , resulting in different genes being more important in one or another condition . Another example of a condition-specific distant peak is provided by the eth11/gxe13 peak on chromosome 15 ( 160–220 cM ) . A total of 386 transcripts show distant linkage to this locus in ethanol , whereas in glucose , only ten show distant linkage . In contrast to most other peaks , the transcripts linking here do not show functional enrichment for gene ontology ( GO ) terms . Targets of similar signaling pathways should share transcription factor binding sites , even if the functional implications of the pathway are unknown . To determine whether we could find enrichment for specific transcription factor binding sites , we split this group of transcripts into four groups , depending on whether they were up or down-regulated by condition and genotype . We used transcription factor binding site data derived from experimental binding as well as species conservation [40] to determine whether any of these gene lists were enriched for experimentally confirmed as well as potential transcription factor binding sites . Those genes that were expressed higher in glucose and in the segregants carrying the RM allele were enriched for sites where MSN2 and MSN4 target sequence was conserved ( MSN2 p = 2 × 10−7 , MSN4 p = 3 × 10−6 ) , but not for sites that have been shown to experimentally bind MSN2 or MSN4 . Experimental binding of MSN2 and MSN4 was tested in a variety of conditions [41] , but these did not include ethanol , and so these targets may reflect an additional functional role for these transcription factors . Those transcripts that show lower expression in ethanol and higher expression in segregants with the BY allele are enriched for CIN5 binding sites that are both conserved and show experimental binding in the conditions tested ( high and moderate hyperoxia , enrichment p = 6 × 10−5 ) . CIN5 is located in this region , shows local linkage with gene–environment interaction , and contains 17 nucleotide changes in the 500 bp upstream of start ( two to three are expected on average , given an upstream polymorphism rate of 0 . 005 [30] ) , in addition to five coding single nucleotide polymorphisms ( SNPs ) ( two nonsynonymous ) , making it an excellent candidate for regulating a subset of these transcripts . CIN5 may not be the regulator for all transcripts linking here—because the peak is very wide , it may contain multiple variants . After examining large differences in distant peaks , we sought to more generally and rigorously describe how the effect of a locus changes across conditions . Comparisons of gxeQTL with linkage results for transcript levels in glucose or ethanol showed that only 965/1 , 555 ( 62% ) gxeQTL also reached genome-wide significance levels in at least one condition ( see Figure S2B for full Venn diagram ) . However , for a test of whether a specific marker that represents a gxeQTL is also linked to the expression level of the corresponding transcript in one of the conditions , setting a genome-wide significance threshold is too conservative . We tested for linkage between the most significant marker at each of the 1 , 555 gxeQTL and expression of the respective transcripts in glucose and in ethanol , lowering the threshold to a FDR of 5% for testing a single marker for each trait , as obtained through inspection of the p-value distribution with QVALUE [42] . All 1 , 555 gxeQTL showed an effect in at least one condition , with 265/1 , 555 ( 17% ) showing an effect only in glucose , 327/1 , 555 ( 21% ) showing an effect only in ethanol , and 963/1 , 555 ( 62% ) showing an effect in both conditions . The fact that all gxeQTL showed linkage to the expression phenotype in at least one condition is not surprising , as a difference in how the genotypes differ between two conditions mathematically requires a change in the mean phenotypic value in at least one condition . The QVALUE method also provides an estimate of the proportion of loci found within a condition that show gene–environment interaction . We inspected the distribution of all p-values for linkage between the most significant marker in one condition and the difference in expression between the conditions , and found that 77% and 76% of loci in glucose and ethanol , respectively , also influence the difference between conditions . Since promoters often define when and where transcripts get expressed , it might be thought that when they do show gene–environment interaction , variants influencing cis-acting regions would be more likely to be condition specific . We did not observe this relationship between local and distant linkages , however . Distant linkages were significantly more likely to be condition specific than local linkages , as 546/1 , 383 ( 40% ) of distant linkages were significant in only one condition , as compared to 46/172 ( 27% ) of local linkages ( Table 2 and see Figure 5 for a schematic ) . Thus , local linkages are overall less dependent on the environment , and when they do show gene–environment interaction , their effects are less likely to be restricted to a specific condition . We investigated whether individual hot spots showed condition specificity by comparing the proportion of gxeQTL in each hotspot that were glucose specific or ethanol specific to the overall rates in all distant gxeQTL using a hypergeometric test ( Figure 6 ) . Five hot spot regions showed altered proportions of condition-specific gxeQTL: peaks gxe2 ( AMN1 ) , gxe5 , gxe6 ( GPA1 ) , and gxe12 ( IRA2 ) were enriched for glucose-specific linkages , whereas peak gxe13 was enriched for ethanol-specific linkages . These results agree with the observation that distant peaks change between conditions , but provide a more quantitative result that indicates that some distant loci show differential effects between conditions . This may be due to changes in the activity states of the proteins and pathways involved as conditions change , or it could also be due to epistasis that masks the effects of the polymorphisms in the alternate condition . Loci showing gene–environment interaction that have effects in both conditions may act by increasing expression in one condition and decreasing expression in the other or by affecting expression in the same direction in both conditions , but to a different extent ( see Figure 5 for a schematic ) . Of those gxeQTL that had an effect in both conditions ( 963/1 , 555 ) , the effects were in opposite directions 72% of the time . This phenomenon was dependent on the type of locus , as local gxeQTL were more likely to act in the same direction in both conditions ( 66% ) , whereas distant gxeQL were more likely to act in opposite directions ( 78% ) ( Table 3 ) . The majority of the loci that had effects in opposite directions did not reach genome-wide significance in either condition ( 463/697 ) . Exclusion of loci that were not genome-wide significant in at least one condition only slightly reduced the association between the type of locus and the direction of effect ( RR = 2 . 3 vs . 3 . 0 ) . The high rate of distant loci acting in opposite directions is not due to an individual peak , as most distant peaks , as well as the set of all linkages that fall outside of peaks , have a high proportion of linkages that act in opposite directions ( Figure 6 ) . This pattern is consistent with the result that local gxeQTL are less likely to show gene–environment interaction overall and less likely to be condition specific . We further investigated peak gxe12 , located at chromosome 15 50–80 cM , which we observed to be enriched for transcripts that show glucose-specific effects . A total of 372 transcripts link here distantly , and many are generally involved in energy metabolism and growth ( Figure 7A ) . Most of these transcripts show large changes between glucose and ethanol , but the allele that the segregant inherits determines the strength of the change . The transcripts that are repressed in ethanol are enriched for ribosomal biogenesis and assembly ( p = 8 × 10−13 ) , whereas those that are activated in ethanol are enriched for generation of precursor metabolites and energy ( p = 1 × 10−6 ) . IRA2 is a strong candidate for the regulation of these transcripts . Ira2 is an inhibitor ( GTPase activating protein ) of the Ras proteins , which mediate the cellular response to glucose via the Ras/PKA pathway [43] . Components of this pathway are highly conserved across species , and Ira2 is a homolog of the neurofibromin tumor suppressor in humans [44] . The IRA2 coding region is highly polymorphic , with 87 SNPs and a single 3-bp indel differing between the strains , resulting in 61 synonymous changes , 26 nonsynonymous changes , and one RM insertion . Of the 26 nonsynonymous changes , BY carries the ancestral amino acid ( Saccharomyces paradoxus ) in 20 cases , and RM carries the ancestral amino acid in four cases , with the remaining two different from S . paradoxus in both strains . To determine whether polymorphism within IRA2 is responsible for the observed linkages , we generated allele replacement strains that carried the coding region of each parent strain in the background of the other parent . We felt that targeting the coding region was appropriate because the expression of IRA2 has not shown local linkage [21] . We expression profiled the replacement strains in glucose and ethanol , and compared the effect of the replacement to the effect of the locus observed in the segregants ( Figure 7B and 7C ) . We observed high correlations between replacements in both parental backgrounds and segregant effects for transcripts showing gene–environment interaction , and the regression slope for both was close to 1 . Many transcripts link to this region in glucose and ethanol ( 1 , 159 and 410 , respectively ) , and IRA2 polymorphisms also contribute to the variation in transcript levels within these conditions ( Figure S4 ) . Thus , polymorphism in IRA2 appears to be the major contributor to the difference in expression between conditions as well as to transcript abundance in each condition for many transcripts linking to this region . This is consistent with coding variants in Ira2 influencing transcript levels via changes in the signaling state of the Ras/PKA pathway . To determine which allele had a stronger effect on the Ras/PKA pathway , we compared the IRA2 alleles to their respective knockout strains . We knocked out IRA2 in both parental strains , grew the resulting strains in glucose and ethanol , and then measured the transcriptional response , as compared to the parental strain ( Figure 8 ) . If the allele is active under the conditions that we tested , then the knockout should look different from the parental strain . We observed that both knockout strains were different from the parental strains , with linear regression slopes less than 1 ( BY slope = 0 . 75 , 95% CI: 0 . 71–0 . 78; RM slope = 0 . 40 , 95% CI: 0 . 35–0 . 45 ) , but the RM knockout showed a greater difference from the parental strain ( Figure 8 ) . Since knocking out IRA2 in RM causes a bigger change than knocking it out in BY , this suggests that the RM allele of IRA2 is better at inhibiting the Ras/PKA pathway in these conditions . The BY allele of IRA2 could be a poorer inhibitor because BY has lost some ability to inhibit or because RM has gained this ability . We took a sequence-based approach and compared IRA2 coding sequences from the three sequenced S . cerevisiae strains ( S288c , isogenic to BY; RM11-1a; and YJM789 [YJM] ) , with S . paradoxus as an outgroup . The S288c sequence is more similar to the ancestral S . paradoxus than either RM or YJM , indicating that RM IRA2 is the more diverged allele . BY is more similar to the ancestral sequence , suggesting that RM has gained the ability to inhibit signaling more . However , since we do not know the specific causal variant ( s ) , we are unable to rule out the possibility that a small number of polymorphisms that are not characteristic of the overall trend are responsible for a decrease in the BY allele's ability to inhibit . If the RM allele has gained the ability to more strongly inhibit Ras/PKA signaling , then it might be expected to show signs of selection . We used the McDonald-Kreitman test to look for evidence of selection in IRA2 by comparing rates of nonsynonymous and synonymous substitutions at polymorphic versus fixed sites within the three sequenced S . cerevisiae strains , as compared to the outgroup S . paradoxus . Of the sites that were fixed in S . cerevisiae , 133/882 ( 15% ) were nonsynonymous , whereas within polymorphic sites , this rate was 2-fold higher ( 31/103 , 30% ) . This difference is significant ( Fisher exact p = 0 . 0001 ) , suggesting that positive selection , or a relaxation of negative selection , has occurred in the S . cerevisiae strains . The RM strain appears to be contributing the most to the nonsynonymous rates , as the ratio of nonsynonymous to synonymous changes along the RM branch is the highest within the strains . Thus , the RM allele appears to have experienced a change in selection pressure in the past . Further analyses of a variety of yeast isolates should help elucidate the evolutionary history of this locus .
Here , we have used yeast gene expression as a model system to describe gene–environment interaction at strain , locus , and gene levels . We showed that 2 , 037 transcripts were jointly dependent on strain and condition in two parental strains . Then , we performed linkage analysis on the difference in transcript levels across conditions with 109 segregants , and identified 1 , 555 gxeQTL . The high number of gxeQTL that we detected has allowed us to make some general observations . We have shown that local and distant linkages differ dramatically in how they act across multiple conditions . Local linkages appear to be more stable: they are less likely to be dependent on the environment , and even when they are , they are more likely to have an effect in both conditions , with the direction of effect often being the same . Distant linkages , on the other hand , are more volatile: they are more likely to be dependent on condition and to show an effect in only one condition . Entire distant peaks can change across conditions , and when they do have an effect in multiple conditions , distant loci are more likely to act in different directions . Finally , we characterized the gene responsible for influencing the largest gene–environment interaction distant peak , IRA2 . We showed that the RM allele of IRA2 is a stronger inhibitor of Ras/PKA signaling than the BY allele in the conditions that we tested , and that this locus has experienced a change in selective pressure in RM . Previous studies have reported that local linkages are more consistent than distant linkages across conditions and experiments , including worms in different temperatures [6] , different tissues in mice [15 , 16 , 33] , and in the reproducibility of transcript linkages in human studies [10 , 34] , indicating that this pattern is likely to extend beyond yeast . Since local and distant linkages are likely to differ in how they influence traits on a molecular level , we can speculate as to how they show differences in condition dependence . Although we do not know most of the causative polymorphisms involved for each type of linkage , local linkages show increased rates of polymorphism in 5′ and 3′ noncoding regions and high rates of allele-specific expression [30] . Thus , we feel comfortable treating the two groups as distinct entities: local linkages are likely to be enriched for variants that directly influence transcript levels via changes in cis-regulatory sites , whereas distant linkages typically influence levels via a protein intermediate ( trans factors ) . In a cis-regulatory site that interacts with a binding protein to directly increase or decrease transcript levels , mutations can either disrupt or enhance binding , but are unlikely to be able to do both in a condition-specific manner . If the binding protein is an activator , a loss of binding mutation will result in lower transcript levels in all conditions where the activator is present , and will show no change when the activator is absent , resulting in either a condition-specific pattern or in a pattern in which the locus has the same effect or the same direction of effect in both conditions ( Figure S5A ) . Other than the case where a single polymorphism destroys a site for one binding factor while creating a site for another , it is less clear how one cis-regulatory mutation could be associated with effects in opposite directions , although one might imagine more complicated scenarios with either multiple linked cis-regulatory polymorphisms or transcription factors that are able to act as both activators and inhibitors . On the other hand , distant variants that influence protein intermediates have the potential to interact with many proteins , depending on the milieu present in the cell in a given condition ( Figure S5B ) . A single variant may be able to activate transcription in one condition and repress it in another , resulting in a change in direction of effect . The frequent occurrence of locus effects in opposite directions in the two conditions is surprising . One possible explanation is multiple linked polymorphisms . Distant loci are much larger targets for variation than cis-regulatory regions [45] , but this could occur in both contexts . Multiple compensatory mutations could accumulate at loci and , depending on the condition , could compensate differentially . The mean phenotype would be stable over conditions , yet the direction of the effect within a condition could vary ( Figure S5C ) . One example is suggested by the gxe11 peak on chromosome 14 , where multiple polymorphisms that influence sporulation [46 , 47] and high-temperature growth [48] have been characterized , and some act in the opposite direction of the overall locus effect . At least one of these alleles ( MKT1 D30G ) is at least partially responsible for the gxeQTL in this region ( Figure S6 ) . Further characterization of the polymorphisms involved at these loci should help elucidate the underlying mechanisms behind this phenomenon . A practical implication of the observation that distant loci often act in opposite directions in the two conditions is that such loci may be inherently difficult to detect in experiments where condition is not controlled , as when different tissues of multicellular organisms are mixed or when nonexperimental organisms ( including humans ) that experience different unmeasured environments are studied . This is because when conditions are not controlled , effects of opposite direction will cancel each other out , resulting in no overall association between the trait and the locus . This also has implications for selection , as these loci may be hidden from selection in organisms that experience fluctuating environmental conditions . The detection of loci that act in opposite direction is additionally complicated by our observation that the majority of these loci did not reach genome-wide significance in either condition alone , emphasizing the importance of using methods to directly test for gene–environment interaction without prior reliance on linkage in a single environment . One general implication of our results for studies in other species , including humans , is that many genetic effects on most traits are likely to be detected without testing for gene–environment interactions , provided that the relevant environmental factors are known and controlled either experimentally or statistically . However , analyses that ignore gene–environment interactions introduce strong biases with regard to the types of loci that are detected . Moreover , gene–environment interactions play a dominant role for a minority of traits . We have studied the prevalence and importance of gene–environment interactions in a single-cell organism grown in two very different and precisely controlled environmental conditions . Our focus on transcript levels as quantitative traits allowed us to study a very large number of traits simultaneously and to delineate general patterns , as well as to provide detailed molecular examples of loci that show gene–environment interactions . The quantitative details would undoubtedly differ if different species , environments , and phenotypes were studied . It is possible that many environmental differences experienced by humans may be less drastic than those between growth on glucose and ethanol are for yeast . However , some environmental differences ( for example , exposure to pathogens or shift from traditional to modern diets ) can have a dramatic effect on health . Our detailed studies in a model organism provide examples of the types of effects that may be expected in humans , and thereby inform practical study design . Understanding the subset of human genetic variants whose phenotypic effects are modifiable by the environment will be key in making full use of personal genomic variation .
The segregants used in this study have previously been expression profiled in condition similar to the glucose condition used here [21] . However , in order to make conditions as similar as possible for this study , with only the carbon source differing , these experiments were repeated alongside the experiments in ethanol . Yeast strains were grown in media consisting of 6 . 7g/l Yeast Nitrogen Base ( Sigma Y0626 ) , 100 mg/l leucine , 100 mg/l lysine , 20 mg/l uracil , and one of the following: 1% ( v/v ) ethanol or 2% ( w/v ) glucose . All growth was performed at 30 °C . After streaking to rich medium ( YPD ) plates from the freezer , cells were pregrown in 5 ml of the phenotyping medium overnight , while spinning on a rotor drum . Approximately 5 × 105 ( glucose ) or 107 ( ethanol ) cells were transferred to 15 ml of fresh medium in a 125-ml Erlenmeyer flask and grown overnight on a rotary shaker at 200 rpm . Cells were harvested at an optical density at 600 nm ( OD600 ) between 0 . 36 and 0 . 40 . Cells were collected ( 5 ml ) at 30 °C by vacuum filtration , and the filters were immediately frozen in liquid nitrogen . Replicates of parental strains and parental strains used to generate the reference sample were grown interspersed with segregant strains such that the parental variation would reflect experimental variation throughout the experiment . RNA was extracted from the filters using a standard hot acid phenol method , followed by RNeasy cleanup ( Qiagen ) . Samples were quality controlled with RNA 6000 Nano kit of the Bioanalyzer 2100 ( Agilent ) . Samples were labeled with either Cy3 or Cy5 dye , using the Low RNA Input Linear Amp Kit ( Agilent ) with the modification that half reactions were used with a quarter of the recommended dye . All samples were hybridized against the same common reference consisting of equal amounts of RNA from both parents ( BY and RM ) grown in both conditions ( glucose and ethanol ) . Six independent replicates of each strain–condition combination were grown and harvested , for a total of 24 samples . RNA was isolated for each sample individually and then pooled , with an equal amount of RNA contributed by each sample . Multiple labeling reactions ( 24 for each dye ) were pooled , and the same samples were used for all segregant and parental arrays . Experiments for knockout and allele replacement arrays used the same reference RNA sample , but a new labeling reaction . Samples were hybridized to Agilent 11k yeast arrays , which are two-color , 60-mer oligo arrays with two arrays per slide , each containing spots for 6 , 256 transcripts , with some duplicated spots . In order to minimize experimentally induced bias , we randomized the order in which samples were processed , hybridizing glucose and ethanol samples at the same time and on the same slides . After hybridizing and washing per Agilent instructions , the arrays were scanned using an Agilent scanner and analyzed with Agilent's Feature Extraction software ( versions 8 . 0–9 . 5 ) . The arrays were uploaded into PUMAdb for processing ( http://puma . princeton . edu/ ) . Spots were considered good data if intensity was well above background and the feature was not a nonuniformity outlier . Additionally , transcripts were retained if data were present in all parental strains in the parental analysis or in at least 80% of the segregants in both conditions for the segregant analysis . We ultimately used 4 , 342 transcripts for the parental analysis and 4 , 482 transcripts for the segregant analysis . Six replicates of each parent ( BY and RM ) in each condition ( glucose and ethanol ) were grown , expression profiled , and labeled ( half with Cy3 and half with Cy5 ) . There were 4 , 342 transcripts with high-quality data: there were no polymorphisms within the probe sequence , and good data was present for each of the replicates . For each transcript , an ANOVA of the form was performed in R using the aov function . In order to assess significance , the dataset was permuted with respect to strain and condition 1 , 000 times , retaining six values of each strain–condition combination , and repeating the ANOVA for each transcript . A FDR was calculated for a range of p-values for strain , condition , and strain–condition interaction effects separately . p-Value cutoffs of 0 . 04 , 0 . 04 , and 0 . 03 corresponded to FDRs of 5% for strain , condition , and strain–condition interaction effects , respectively . Genetic correlation was calculated from the variance results of the ANOVA study using the following equation [49] . Segregants were grown and expression profiled in both conditions , resulting in a total of 218 segregant arrays: 109 segregants each in glucose and ethanol . Samples were randomized with respect to dye , and all arrays were linearly adjusted for dye . Three phenotypes were used in linkage analysis: expression in glucose , expression in ethanol , and the difference in expression between glucose and ethanol ( ethanol minus glucose for each segregant ) . Mapping a significant locus for the change across conditions is interpreted as gene–environment interaction . It can also be conceptualized as a paired test where the segregant's ethanol phenotype is effectively corrected for its phenotype in glucose . Genotypes have been previously characterized [21] , but we further refined the map by removing a set of 62 markers , corresponding to 13 Affymetrix gene regions , that were artificially inflating the map . When each gene region was removed , the total map distance ( Haldane ) decreased at least 25 cM . All but one of these gene regions corresponded to sequence that was duplicated in the genome , which would increase genotyping error . Our final set of markers consisted of 2 , 894 markers with an average spacing of 4 kb or 1 . 9 cM . Linkage analysis was performed using the nonparametric method of the qtl package in R [50] , which is an adaptation from the Kruskal-Wallace test and is similar to that described by Kruglyak and Lander [51] . In addition to the genotypes of the markers , genotype probabilities were estimated at every centimorgan along the genome using the calc . genoprob function . Significance was assessed by permutation . For each permutation , segregants were assigned to a new array randomly , and linkage analysis on all transcripts was repeated . The permutation was performed ten times and the average number of transcripts showing linkage at a specific logarithm of the odds ( to the base 10 ) ( LOD ) score was used to calculate a FDR . LOD scores of 3 . 25 , 3 . 3 , and 3 . 7 corresponded to FDRs of 4 . 9% , 4 . 9% , and 4 . 8% for glucose , ethanol , and the difference between condition phenotypes , respectively . We also calculated confidence intervals as LOD support regions that extend outward from the peak marker until the LOD score decreases by 1 . 5 LOD units [52] . A linkage was classified as a local linkage if the confidence interval of the linkage overlapped with the region containing the open reading frame of the gene of interest ( ORF region ) . The two markers flanking the region from 500 bp upstream of the start of the ORF to 500 bp downstream of the stop of the ORF , as defined by the Saccharomyces Genome Database , defined the ORF region . A distant linkage is any linkage that fell outside of the ORF region . The genome was broken into 10-cM bins , and the peak of each linkage in each condition ( glucose , ethanol , or gene–environment interaction ) was assigned to a bin . A bin was considered to have an excess of linkages if the number exceeded the number expected by chance by Poisson distribution , given the total number of distant linkages and Bonferroni correction for 563 tests ( p < 9 × 10−5 ) . These cutoffs were greater than 15 for glucose , 14 for ethanol , and nine for interaction . Significant bins that were located immediately next to each other were merged into a single peak . Peaks in different conditions were considered overlapping if any bin contained in one peak was contained in the other . For each gxeQTL , we took the marker closest to the linkage peak and computed the LOD score for linkage of this marker to the transcript levels in glucose and in ethanol . We then converted the LOD score to a nominal p-value by comparison to permuted data . We created 1 , 000 random phenotypes and performed nonparametric linkage analysis on each . We translated LOD scores to p-values for a given marker by counting the proportion of the randomized phenotypes that had a LOD score as high or higher than that observed . By inspecting the p-value distribution of gxeQTL marker linkages in each condition using QVALUE software [42] , we determined cutoffs that allowed us to call individual linkages significant at a FDR of 5% ( p = 0 . 228 in ethanol and p = 0 . 22 in glucose ) . To estimate the proportion of linkages within a condition that also showed gene–environment interaction , we calculated p-values for the linkage between peak markers found within a condition and the difference in expression between conditions . The proportion of significant tests is . RR estimates and 95% CIs were calculated in R using the twoby2 command of the Epi package . Values of χ2 and p-values were calculated in Excel with expected values derived from row and column expectations . The Mann-Whitney test was performed in R using the wilcox . test command of the stats package . The RM genome was downloaded from the Broad Web site and aligned to the S288c genome ( SGD ) . The sequence of each probe was found in the S288c sequence , and the corresponding sequence determined for RM . We were able to find 6 , 217 probes in the alignment , and of these , 5 , 029 have no polymorphisms at all , 747 have a single polymorphism or gap , and 438 have two or more strain differences . Missing probes were either in the mitochondrial genome or differed from the most recent reference sequence . The presence of even a single polymorphism was associated with an increased probability of apparent local linkage , as would be expected if there were differences in binding efficiencies of the alleles . We thus excluded them from all further analysis . IRA2 and DIG1 replacement strains were generated by a two-step allele replacement method [53] . For example , IRA2 was replaced with URA3 in BY4724 [54] ( MATa ura3 lys2 ) and RM11-1a [9] ( MATa leu2 ura3 ho::KAN ) , generating ira2Δ::URA3 knockout strains . New IRA2 alleles were amplified by PCR with approximately 500-bp overlapping sequence and introduced into the appropriate background to replace URA3 . See Table S2 for strain descriptions . Allele replacements were sequenced to ensure that no new mutations were introduced . For both replacements , the entire coding sequence was exchanged , but the extent that the 3′UTR polymorphisms were exchanged varied . Adam Deutschbauer kindly provided the MKT1 D30G replacement strain in the BY4742 background [46] . The RM IRA2 sequence was obtained from the whole-genome sequencing project at the Broad Institute ( http://www . broad . mit . edu/annotation/genome/saccharomyces_cerevisiae/Home . html ) , with the exception of a small gap , which we sequenced using standard dideoxy methods . Replacement strains were also sequenced using standard dideoxy sequencing methods . We quantified how well the IRA2 replacement strains recapitulated the effect due to the locus in the segregants by comparing the locus effect to the replacement effect for all linking transcripts . For a given transcript , the locus effect is the difference between the BY change across condition ( average of all segregants carrying the BY allele in ethanol minus the average of all segregants carrying the BY allele in glucose ) and the RM change across condition ( average of all segregants carrying the RM allele in ethanol minus the average of all segregants carrying the RM allele in glucose ) . For the replacement effect , the calculation is similar , except instead of segregant strains , we use the parental and replacement strains . Thus , for a single transcript , the replacement effect would be: We then compared the locus effect to the replacement effect across all transcripts by calculating a Pearson correlation . We estimated significance of this correlation by randomizing the genotype at the marker nearest to IRA2 and repeating the analysis 1 , 000 times . The p-value is the number of times that we got a correlation as high or higher than what was observed . A significant correlation indicates that gene that was replaced functionally influences the traits . We also analyzed this relationship by linear regression using the lm command in R . If the regression slope is close to 1 , then it indicates that the gene is a major contributor to the phenotype . Of the 4 , 482 transcripts that were analyzed in the segregants , 4 , 339 had GO terms associated with them . This set was used as a background for GO term enrichment . Hypergeometric tests were performed using GOLEM [55] , and p-values were Bonferroni corrected for multiple testing . Of the 4 , 482 transcripts that were analyzed in the segregants , 4 , 465 were used to look for enrichment of transcription factor ( TF ) binding sites . All nine “ORFs_by_factor” files that are provided by MacIsaac et al . on their Web site ( available at: http://rd . plos . org/pbio . 0060083 ) were used to look for enrichment of binding sites . These lists are constructed by creating lists of genes that contain a predicted TF site at three levels of conservation and three levels of experimental binding [40] . In order to be in a list , the site has to occur within the upstream intergenic region of the gene ( K . MacIsaac , personal communication ) . Hypergeometric tests of enrichment were performed in GOLEM using modified input files in which each gene list corresponded to its own term . p-Values were Bonferroni corrected . The McDonald-Kreitman test [56] was performed using the libsequence library's MK test [57] with sequences from RM , BY , YJM789 [58] , and S . paradoxus [59] . Sequence trees and estimates of nonsynonymous to synonymous rates on individual branches were calculated in HYPHY [60] .
Gene accession numbers reference the Saccharomyces Genome Database ( http://www . yeastgenome . org ) : AMN1 ( S000000362 ) , CIN5 ( S000005554 , GPA1 ( S000001047 ) , HAP1 ( S000004246 ) , HXT6 ( S000002751 ) , HXT7 ( S000002750 ) , IDP2 ( S000004164 ) , IRA2 ( S000005441 , MATALPHA1 ( S000000636 ) , MSN2 ( S000004640 ) , and MSN4 ( S000001545 ) . Expression raw data can be obtained via Gene Expression Omnibus ( GEO; http://www . ncbi . nlm . nih . gov/projects/geo/ ) accession number GSE9376 and through PUMAdb ( http://puma . princeton . edu/ ) . | Individuals frequently encounter different environmental conditions , and the physiological and behavioral responses to these conditions can depend on an individual's genetic makeup . This phenomenon is known as gene–environment interaction . For example , individuals who are infected with the Plasmodium falciparum parasite are susceptible to malaria , but not if they carry the sickle-cell allele of hemoglobin . The general properties of gene–environment interaction are poorly understood , and a better understanding is essential if individuals are to make informed health choices guided by their genomic information . We have investigated gene–environment interaction on a genomic level , characterizing its role in over 4 , 000 traits at once by investigating natural variation in yeast gene expression . We compared lab and vineyard strains of yeast growing in two conditions ( glucose and ethanol as carbon sources ) in which they adopt two different metabolic states: fermentation and aerobic respiration , respectively . We show that gene–environment interaction is a common phenomenon , describe how different classes of genetic variants affect the nature of the interactions , and provide detailed molecular examples of interactions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"genetics",
"and",
"genomics"
] | 2008 | Gene–Environment Interaction in Yeast Gene Expression |
Sfl1p and Sfl2p are two homologous heat shock factor-type transcriptional regulators that antagonistically control morphogenesis in Candida albicans , while being required for full pathogenesis and virulence . To understand how Sfl1p and Sfl2p exert their function , we combined genome-wide location and expression analyses to reveal their transcriptional targets in vivo together with the associated changes of the C . albicans transcriptome . We show that Sfl1p and Sfl2p bind to the promoter of at least 113 common targets through divergent binding motifs and modulate directly the expression of key transcriptional regulators of C . albicans morphogenesis and/or virulence . Surprisingly , we found that Sfl2p additionally binds to the promoter of 75 specific targets , including a high proportion of hyphal-specific genes ( HSGs; HWP1 , HYR1 , ECE1 , others ) , revealing a direct link between Sfl2p and hyphal development . Data mining pointed to a regulatory network in which Sfl1p and Sfl2p act as both transcriptional activators and repressors . Sfl1p directly represses the expression of positive regulators of hyphal growth ( BRG1 , UME6 , TEC1 , SFL2 ) , while upregulating both yeast form-associated genes ( RME1 , RHD1 , YWP1 ) and repressors of morphogenesis ( SSN6 , NRG1 ) . On the other hand , Sfl2p directly upregulates HSGs and activators of hyphal growth ( UME6 , TEC1 ) , while downregulating yeast form-associated genes and repressors of morphogenesis ( NRG1 , RFG1 , SFL1 ) . Using genetic interaction analyses , we provide further evidences that Sfl1p and Sfl2p antagonistically control C . albicans morphogenesis through direct modulation of the expression of important regulators of hyphal growth . Bioinformatic analyses suggest that binding of Sfl1p and Sfl2p to their targets occurs with the co-binding of Efg1p and/or Ndt80p . We show , indeed , that Sfl1p and Sfl2p targets are bound by Efg1p and that both Sfl1p and Sfl2p associate in vivo with Efg1p . Taken together , our data suggest that Sfl1p and Sfl2p act as central “switch on/off” proteins to coordinate the regulation of C . albicans morphogenesis .
Candida albicans is the most frequent causative agent of superficial as well as disseminated , life-threatening fungal infections [1] . The success of C . albicans as a major fungal pathogen of humans relies on a number of pathogenic traits , among which its capacity to grow and switch between at least three distinctive morphological forms: budding yeast , pseudohyphae and hyphae [2]–[5] . The morphogenetic transition has been commonly described as a critical trait for survival and virulence in the host , even though the analysis of a wide array of C . albicans knock-out mutants suggests that pathogenesis can be dissociated to some extent from morphological switching [6]–[8] . The yeast-to-hyphae transition is triggered by a variety of environmental stimuli including nutrient availability , temperature , pH , CO2 and serum [9]–[13] . This process correlates with the coordinated expression of a set of hyphal-specific genes ( HSGs ) with roles in orchestrating hyphal development . Consequently , the transition is highly regulated and involves multiple interconnected signalling pathways , including the cyclic AMP-dependent Protein Kinase A ( cAMP-PKA , regarded as playing a central role in the control of morphogenesis ) , the Cph1p-mediated Mitogen-Activated Protein Kinase ( MAPK ) and the Rim101p-mediated pH cascade pathways , all of which positively regulate hyphal development through the modulation of the activity of transcription factors to control the expression of HSGs ( see [13] for a recent review ) . These transcription factors include ( among others ) Efg1p/Flo8p , acting downstream of cAMP-PKA [14]–[20] , Tec1p [21] and Ume6p [22] , [23] . Hyphal morphogenesis is also subject to negative regulation mostly by the general corepressor Tup1p through interaction with the transcriptional repressors Nrg1p and Rfg1p [4] , [12] , [24]–[27] . In the yeast Saccharomyces cerevisiae , which has been used as a model for studying the transcriptional control of the morphological transition [28] , [29] , Sfl1p ( ScSfl1p , for suppressor gene for flocculation 1 ) is a target of the cAMP-PKA pathway [30] . ScSFL1 encodes a negative regulator of pseudohyphal growth and invasion [31] and was isolated based on its ability to suppress flocculation defects in yeast [32] . ScSfl1p carries a putative heat shock factor ( HSF ) -type DNA binding domain and binds in vitro to a GAA triplet motif [33] characteristic of heat shock elements ( HSEs ) [34] , while exerting its negative regulation through the recruitment of the Ssn6p-Tup1p corepressor complex [35] . ScSfl1p has dual activator/repressor functions , acting as a transcriptional repressor of flocculation-related genes and as an activator of stress-responsive genes [35] , [36] . Interestingly , the C . albicans genome encodes two structural homologs of ScSfl1p , namely Sfl1p and Sfl2p [37]–[40] . Either SFL1 or SFL2 functionally complement an S . cerevisiae sfl1 mutation [38] , [39] and encode important regulators of morphogenesis and virulence in C . albicans [37]–[40] . Intriguingly , although sharing structural homologies , Sfl1p and Sfl2p have antagonistic functions: while Sfl1p acts as a negative regulator of hyphal development , Sfl2p acts as a positive regulator of this process [37]–[40] . Functional analyses of C . albicans Sfl1p showed that deletion of SFL1 promoted filamentous growth and cell flocculation and correlated with induction of HSGs ( ECE1 , HWP1 ) and genes involved in cell adhesion ( ALS1 , ALS3 ) , whereas its overexpression inhibited hyphal formation [37] , [38] . Consistent with a transcriptional regulatory function , an Sfl1p-GFP fusion localized to the nucleus , while one hybrid lacZ reporter analyses in C . albicans correlated with a repressor function [37] . Importantly , either deletion or overexpression of SFL1 attenuated C . albicans virulence in a mouse model of systemic infection [38] . On the other hand , we and others have shown that deletion of SFL2 impaired filamentation in response to different cues , whereas SFL2 overexpression promoted hyphal growth , even under non hyphae-stimulating conditions [39]–[41] . Noteworthy , an sfl2Δ/sfl2Δ strain exhibited reduced damage in a reconstituted human oral epithelium model and displayed attenuated virulence in a mouse model of gastrointestinal colonization and dissemination model [39] , [40] , indicating that Sfl2p also plays an important role in C . albicans pathogenesis . Similar to Sfl1p , an Sfl2p-GFP fusion localized to the nucleus , in line with a role in transcriptional regulation [39] . It is still unknown how Sfl1p and Sfl2p exert their antagonistic functions . Both SFL1 and SFL2 were shown to genetically interact with at least transcription factor FLO8 . Hyphal development in sfl1Δ/sfl1Δ was abolished upon deletion of FLO8 but enhanced upon FLO8 overexpression [38] while overexpression of SFL2 triggered filamentation in a FLO8- and EFG1-dependent manner [39] , suggesting the implication of the cAMP-PKA pathway . It was also shown that SFL2 is required for hyphal maintenance at high temperature and that a temperature increase from 25°C to 37°C leads to upregulation of both the RNA and protein levels of Sfl2p , indicating that Sfl2p is a temperature-responsive regulator [39] . In contrast , no clear association was determined between temperature and Sfl1p function . Interestingly , Song et al . showed that the putative HSF domains of Sfl1p and Sfl2p were required for their functional divergence by testing HSF domain-swapped hybrids for their ability to retain their effect on filamentation [39] . This suggests that the two putative HSF domains in Sfl1p and Sfl2p mediate the specific recognition of divergent target sites that determine the activation or repression roles of Sfl1p and Sfl2p [39] . To shed more light on Sfl1p and Sfl2p functions , we provide a comprehensive functional portrait of these two regulators using a combination of genome-wide location , genome-wide expression and genetic interaction analyses . We provide evidences that Sfl1p and Sfl2p act as central “switch on-off” proteins to coordinate the regulation of C . albicans morphogenesis and , potentially , pathogenesis and virulence .
To better characterize the function of Sfl1p and Sfl2p , we sought to identify their DNA-binding locations , in vivo , by chromatin immunoprecipitation . To this end , we generated triple-hemagglutinin epitope ( HA3 ) -tagged versions of SFL1 and SFL2 and used the pCaEXP system [42] to drive MET3 promoter-dependent expression of the tagged alleles in sfl1Δ/sfl1Δ ( Table 1; strain sfl1-CaEXP-SFL1-HA3 ) and sfl2Δ/sfl2Δ ( Table 1 , strain sfl2-CaEXP-SFL2-HA3 ) mutant strains , respectively ( Figure 1A , see Materials and Methods for specific details ) . We also generated sfl1Δ/sfl1Δ and sfl2Δ/sfl2Δ mutants carrying the empty pCaEXP vector ( sfl1-CaEXP and sfl2-CaEXP , respectively , see Table 1 ) to serve as negative controls for immunoprecipitation . Western blot analyses of strains grown under PMET3-inducing conditions showed that both Sfl1p-HA3 and Sfl2p-HA3 fusion proteins were expressed ( Figure 1B , lanes 4 and 6 ) . As an additional control for signal specificity , immunoblotting of total extracts from a C . albicans strain expressing the Cap1p-HA3 fusion ( Figure 1 , lane 2 ) or the corresponding empty-vector negative control ( Figure 1 , lane 1 ) was used [43] . To test the functionality of the Sfl1p-HA3 and Sfl2p-HA3 fusions , both tagged and empty-vector control strains were grown overnight at 30°C in YPD then transferred to Lee's medium ( hyphae-inducing medium ) lacking methionine ( PMET3-inducing condition ) at 37°C and allowed to resume growth during 4 h prior to microscopic examination ( Figure 1C ) . It was previously shown that PMET3-driven expression of wild-type SFL1 in a homozygous sfl1 mutant strain under hyphae-inducing conditions abolished hyphal formation [37] . As expected , hyphal formation was induced in either the control strain SC5314 or the sfl1Δ/sfl1Δ mutant carrying the empty vector ( Figure 1C , top left and middle panels , respectively ) . Conversely , hyphal formation was strongly impaired in the strain expressing SFL1-HA3 ( Figure 1C , top right panel ) , therefore phenocopying the effect of PMET3-driven wild-type SFL1 expression as observed in Bauer et al . [37] . Under the same growth conditions the sfl2Δ/sfl2Δ strain carrying the empty vector was unable to form hyphae ( Figure 1C , bottom middle panel ) , whereas expression of the SFL2-HA3 allele allowed induction of hyphal formation as observed in strain SC5314 ( Figure 1 , compare bottom left and right panels ) . Taken together , these results show that epitope-tagging of Sfl1p and Sfl2p at their C-termini using the pCaEXP system allowed the production of fully functional proteins . We performed genome-wide location of Sfl1p or Sfl2p under hyphae-inducing conditions by chromatin immunoprecipitation coupled to massively parallel high-throughput sequencing ( ChIP-Seq , see Materials and Methods ) , which allows to detect binding events at a single nucleotide resolution . The resulting reads were mapped to the C . albicans Assembly 21 genome and alignments were visualized using the Integrative Genomics Viewer ( IGV ) software [44] , [45] ( see Materials and Methods for details ) . Using the Model-Based Analysis for ChIP-Seq ( MACS ) peak-finding algorithm [46] , we identified 163 and 213 binding peaks for Sfl1p and Sfl2p , respectively ( see Tables S1–S6 in Text S1 , Legends to Supplementary Tables S1–S8 in Text S1 and Materials and Methods for details ) . As expected , most of Sfl1p or Sfl2p binding peaks were located at ‘intergenic’ regions ( Tables S1–S6 in Text S1 ) , consistent with a transcriptional regulatory function . Among the 163 Sfl1p binding peaks , 76 clearly associated with individual ORFs , while 34 were located at promoter regions shared by two ORFs in opposite orientations and the remaining 53 peaks were not clearly associated with ORFs . In particular , spurious binding overlapping with highly transcribed regions [47] , mostly tRNA-encoding genes , or regions with repeated DNA sequence ( Table S3 in Text S1 ) , was observed . Among the 213 Sfl2p binding peaks , 140 clearly associated with unique ORFs , while 54 were located in promoter regions shared by two ORFs in opposite orientations and the remaining 19 peaks were not clearly linked to defined ORFs ( Table S6 in Text S1 ) . Additional bona fide Sfl1p ( 14 peaks ) and Sfl2p ( 28 peaks ) binding peaks were not detected by the peak-finding algorithm and were added to our target lists ( Tables S3 and S6 in Text S1 , see column entitled “comments” and Legends to Supplementary Tables S1–S8 in Text S1 ) . Overall , examination of Sfl1p and Sfl2p binding peaks allowed to identify 113 and 188 target promoters ( Figure 1A ) including 39 and 56 promoter regions shared by two ORFs , respectively . Interestingly , all 113 Sfl1p targets were also bound by Sfl2p , suggesting functional interactions between the two regulators , while 75 additional targets were specific to Sfl2p ( Figure 2A ) . In many occurrences , Sfl2p binding at promoter regions strongly overlapped with that of Sfl1p ( Figure 2B , top panel as an example ) . In other cases , Sfl2p binding showed partial ( Figure 2B , middle panel as an example ) or no overlap ( Figure 2B , bottom panel as an example ) with Sfl1p binding . Noteworthy , Sfl2p and Sfl1p binding peaks were often lying across relatively long regions , particularly in the vicinity of transcription factor-encoding genes such as EFG1 ( Figure 2B , top panel ) , UME6 , NRG1 or TEC1 , suggesting the presence of more than one binding site or the existence of functional interactions with other regulatory proteins at these sites . We used the GO Term Finder tool from the CGD [48] to identify functional enrichment among Sfl1p and Sfl2p targets relative to the annotated C . albicans genome ( Table 2; see Materials and Methods ) . Strikingly , we found that the most significantly enriched functional category among Sfl1p and Sfl2p common targets was “Sequence-specific DNA-binding transcription activity” ( 21 genes , P = 1 . 42×10−8; Figure 2C , grey shading ) , including a large number of genes encoding major transcription factors involved in C . albicans morphogenesis and virulence such as UME6 , TEC1 , NRG1 , RFG1 , BRG1 , FLO8 , others ( Figure 2C and Table 2 ) . In line with this finding , the functional grouping “Filamentous growth” ( 30 genes , P = 1 . 83×10−6 ) was also among the most overrepresented categories of the identified GO terms and included the above-mentioned transcription factors in addition to HMS1 , encoding a transcription factor that controls C . albicans morphogenesis mediated by HSP90 compromise or high temperature [49] , as well as many genes encoding effectors or signal transducers of this process such as MSB2 , CHT2 , GAP1 , ALS1 , RAS2 , others ( Figure 2C ) . As expected , “Pathogenesis” ( 16 genes; P = 2 . 40×10−4 ) was also among the most significantly enriched functional categories among Sfl1p and Sfl2p common targets and is consistent with the known roles of Sfl1p and Sfl2p in C . albicans virulence [38] , [39] . Interestingly , Sfl1p and Sfl2p bound to genes encoding transcription factors involved in white/opaque switching , including WOR2 , FLO8 , EFG1 and AHR1 ( “Regulation of phenotypic switching”; 4 genes; P = 4 . 34×10−2 ) , as well as genes involved in biofilm formation ( “Biofilm formation”; 12 genes; P = 6 . 40×10−4 ) , suggesting wider functions for these two regulators in C . albicans . These functions may include the ability to respond to a variety of stimuli , such as drug treatment ( “Cellular response to drug”; 17 genes; P = 2 . 48×10−3 ) , nutrient availability ( “Cellular response to nutrient levels”; 18 genes; P = 3 . 00×10−3 and “Galactose catabolic process via UDP-Galactose”; 3 genes; P = 2 . 23×10−3 ) and pH levels ( “Cellular response to pH; 9 genes; P = 3 . 62×10−3 ) . We also performed functional category enrichment analyses of the 75 Sfl2p-specific targets ( Figure 2C , unshaded area ) . Interestingly , these targets were grouped into functional categories pertaining to interaction with the host , including “Multi-organism process” ( 19 genes; P = 2 . 06×10−5 ) , “Symbiosis , encompassing mutualism through parasitism” ( 9 genes; P = 2 . 18×10−3 ) , “Adhesion to host” ( 6 genes; P = 2 . 69×10−3 ) and “Fungal-type cell wall” ( 11 genes; P = 1 . 92×10−5 ) . Sfl2p also bound specifically to 11 genes encoding transcription factors such as CPH2 , ECM22 , CZF1 , FCR3 , RFX2 and ROB1 ( Table 2 ) . We also found that Sfl2p bound specifically to the SFL1 promoter , while both Sfl1p and Sfl2p bound to the promoter of SFL2 , suggesting an autoregulatory loop controlling SFL2 expression . To validate our ChIP-Seq data , we performed additional independent ChIP experiments and measured Sfl1p and Sfl2p binding by PCR ( ChIP-PCR ) on selected targets ( Figure 3 ) . The URA3 and YAK1 genes were used as negative controls for ChIP enrichment . As expected , Sfl1p and Sfl2p binding was detected at the promoter of their targets , including BRG1 , EFG1 , SFL2 , UME6 and TEC1 ( Figure 3 ) . The promoter region of Sfl2p-specific targets was also enriched by Sfl2p-HA3 immunoprecipitation , including SFL1 , RBT1 and FAV2 , but not by the immunoprecipitation of Sfl1p-HA3 ( Figure 3 ) . Taken together , our results suggest that Sfl1p and Sfl2p regulate C . albicans morphogenesis and potentially confer virulence through direct binding to the promoter of genes encoding key regulators of these processes . They also revealed that , while both transcription factors bind to common targets , Sfl2p specifically binds to additional target genes that appear to be involved in processes pertaining to interaction with the host . To determine whether Sfl1p and Sfl2p binding targets were also transcriptionally modulated , we performed global gene expression analyses of strains sfl1-CaEXP-SFL1-HA3 versus sfl1-CaEXP and sfl2-CaEXP-SFL2-HA3 versus sfl2-CaEXP grown 3 times independently under the same conditions than those in the ChIP-Seq experiments ( see Materials and Methods for details ) . We found 643 upregulated and 579 downregulated genes ( expression fold-change ≥1 . 5; P≤0 . 05 ) in the sfl1-CaEXP-SFL1-HA3 strain as compared to strain sfl1-CaEXP ( Table S7 in Text S1 ) . On the other hand , 354 genes were upregulated and 478 genes were downregulated ( expression fold-change ≥1 . 5; P≤0 . 05 ) in strain sfl2-CaEXP-SFL2-HA3 relative to sfl2-CaEXP ( Table S8 in Text S1 ) . Data were visualized using an expression profile plot ( GeneSpring version 12 , Agilent Technologies ) , which allows to get a global view of gene expression variation and thus to compare the expression patterns in SFL1 and SFL2 data sets ( Figure 4A ) . Interestingly , most of the highly upregulated genes in pCaEXP-SFL1-HA3 vs . pCaEXP data were strongly downregulated in pCaEXP-SFL2-HA3 vs . pCaEXP data ( Figures 4A and 4B left panel ) . Many of these genes are markers of the yeast form growth phase , such as RME1 , YWP1 , RHD1 and orf19 . 557 . On the other hand , most of the strongly downregulated genes in pCaEXP-SFL1-HA3 vs . pCaEXP data were actually upregulated in pCaEXP-SFL2-HA3 vs . pCaEXP data ( Figure 4A ) , including the HSGs ECE1 , ALS3 , IHD1 , HWP1 , HYR1 and SAP5 ( Figure 4B ) . Examination of the genes that were strongly modulated in pCaEXP-SFL2-HA3 vs . pCaEXP data also revealed similar gene expression dynamics: many of the upregulated genes were found to be downregulated in pCaEXP-SFL1-HA3 vs . pCaEXP data sets , and vice versa ( Figure 4B , right panel ) . We independently confirmed the microarray data by qRT-PCR analyses of selected genes using homozygous sfl1 or sfl2 mutant strains expressing ( or not ) functional TAP ( tandem affinity purification ) -tagged SFL1 or SFL2 alleles [41] , respectively , under the control of the PCK1 promoter ( Figure 5 , Table 1 ) . Strains were grown under gluconeogenic ( PPCK1-inducing ) conditions during 0 , 2 and 4 hours and total RNA was isolated followed by qRT-PCR ( See Materials and Methods for details ) . As expected , expression of SFL1-TAP gradually increased from time points 0 h to 4 h ( Figure 5A , left panel ) . This increased SFL1 expression correlated with decreased SFL2 and BRG1 expression ( Figure 5A , middle and right panels ) , consistent with a negative regulation of SFL2 and BRG1 expression . On the other hand , PPCK1-induced SFL2-TAP expression ( Figure 5B , left panel ) correlated with decreased expression of SFL1 ( Figure 5B , SFL1 panel ) and increased expression of UME6 and ALS3 ( Figure 5B , UME6 and ALS3 panels ) , consistent with our microarray data ( Figure 4 ) . Taken together , our transcriptomics data reflect the antagonistic functions of Sfl1p and Sfl2p in regulating C . albicans morphogenesis , with SFL1 promoting the yeast-form growth which correlates with upregulation of yeast form-specific genes and downregulation of HSGs , and SFL2 promoting hyphal growth which correlates with upregulation of HSGs and downregulation of yeast form-specific genes . We combined the transcriptomics and the ChIP-Seq data in order to get a genome-wide view of the transcriptional modules associated with Sfl1p and Sfl2p regulatory functions ( Figure 6 ) . We were expecting to find a substantial amount of genes that are bound by Sfl1p and downregulated in pCaEXP-SFL1-HA3 vs . pCaEXP microarray data , as Sfl1p is thought to act as a repressor . In line with the function of Sfl2p as an activator of hyphal growth , we were also hypothesizing that binding of Sfl2p to its targets would correlate with increased expression of these target genes . Surprisingly , among the 113 targets commonly bound by Sfl1p and Sfl2p , 40 genes were upregulated and only 22 genes were downregulated in pCaEXP-SFL1-HA3 vs . pCaEXP data ( Figure 6A ) . Conversely , 39 genes were downregulated in pCaEXP-SFL2-HA3 vs . pCaEXP data and only 15 genes were upregulated ( Figure 6A ) , indicating that Sfl1p and Sfl2p have dual transcriptional regulatory functions; acting as both transcriptional activators and transcriptional repressors . As Sfl1p and Sfl2p respectively act as a repressor and an activator of hyphal growth , we examined the set of genes that were commonly bound by these two regulators and whose expression was both downregulated by SFL1 and upregulated by SFL2 . We found 9 genes matching these criteria ( Figure 6A , middle right box ) , among which the key regulators of hyphal growth UME6 and TEC1 . We also examined the set of genes that were both bound by Sfl1p and Sfl2p and upregulated in pCaEXP-SFL1-HA3 vs . pCaEXP and/or downregulated in pCaEXP-SFL2-HA3 vs . pCaEXP microarray data ( Figure 6A , left boxes ) . This is consistent with Sfl1p acting as a transcriptional activator for these genes and/or Sfl2p functioning as their transcriptional repressor . Interestingly , we found that many of these genes encode ( or are predicted to encode , e . g . orf19 . 6874 ) negative regulators of hyphal growth , including SSN6 , orf19 . 6874 [50] , NRG1 and RFG1 ( Figure 6A , left boxes ) . Of particular interest , EFG1 , the major regulator of C . albicans morphogenesis that functions as both a transcriptional activator and a repressor depending on the growth condition [51] was found to be upregulated by Sfl1p but not modulated in SFL2 microarray data . Sfl1p and Sfl2p also bound to the promoter of BRG1 , AHR1 , HMS1 and SFL2 ( Figure 6A ) , all encoding transcriptional activators of hyphal growth . The expression of BRG1 and AHR1 was downregulated by Sfl1p but not modulated by Sfl2p ( Figure 6A , bottom right box ) , whereas the expression of HMS1 was downregulated by Sfl2p but not modulated by Sfl1p ( Figure 6A , bottom left box ) . Interestingly , Sfl1p binding to the SFL2 promoter correlates with decreased expression of SFL2 , indicating a direct negative regulation of SFL2 expression by Sfl1p ( Figures 5A and 6A ) . Sfl2p binding to its 75 specific target genes correlated with increased and decreased expression of 24 and 25 genes , respectively ( Figure 6B ) . Strikingly , a significant subset of the genes that are both bound and transcriptionally induced by Sfl2p were the HSGs ALS3 , HGC1 , HWP1 , HYR1 , ECE1 , SAP4 , IHD1 , FAV2 and RBT4 in addition to DCK1 encoding a putative guanine nucleotide exchange factor required for filamentous growth and the hyphal induced gene orf19 . 3475 ( Figure 6B , upper box ) . Moreover , Sfl2p directly upregulated genes encoding ( or predicted to encode ) transcription factors , including FCR3 , encoding a positive regulator of C . albicans adherence [52] , orf19 . 217 , encoding a positive regulator of hyphal growth [41] and RFX2 , encoding a regulator of DNA damage response , adhesion and virulence [53] . On the other hand , Sfl2p directly downregulated the expression of transcription factors SFL1 , ECM22 , ROB1 , encoding a regulator of biofilm formation [54] , and many genes involved or predicted to be involved in cell wall integrity ( EAP1 , FUN31 , SIM1 , PIR1 and RHD3 ) as well as genes encoding or predicted to encode permeases or transporters ( PHO86 , putative inorganic phosphate transporter; HGT1 , high-affinity glucose transporter; FLC3 , putative heme transporter; HIP1 and orf19 . 7566 , putative amino acid transporters ) . Taken together , combination of the ChIP-Seq and the transcriptomics data i ) indicate that Sfl1p and Sfl2p have dual transcriptional regulatory functions , acting as both activators and repressors , ii ) suggest that Sfl1p and Sfl2p antagonistic functions in regulating hyphal morphogenesis is mediated through direct transcriptional modulation of genes encoding key regulators of C . albicans morphogenesis , iii ) show that Sfl2p additionally specifically controls the expression of HSGs and iv ) reveal a direct SFL1-SFL2 cross-factor negative control . Our finding that Sfl1p and Sfl2p directly control the expression of master regulators of C . albicans morphogenesis and virulence fostered us to assess the genetic interactions between SFL1 , SFL2 and these target genes . Data mining of our ChIP-Seq and transcriptomics results showed that Sfl1p directly negatively regulates SFL2 expression ( Figures 3 , 5A and 6A ) . Moreover , Sfl1p directly negatively regulates the expression of BRG1 ( Figures 3 , 5A and 6A ) , encoding a major regulator of hyphal growth . This suggests that SFL1 represses filamentation through , at least , direct transcriptional repression of the SFL2 and BRG1 genes . To test this hypothesis , we constructed sfl1Δ/sfl1Δ , sfl2Δ/sfl2Δ and sfl1Δ/sfl1Δ , brg1Δ/brg1Δ double mutants and tested their ability to form hyphae ( Figure 7A ) . All strains displayed yeast-form growth in SD medium at 30°C ( Figure 7A , upper panels ) . In YP 10% FBS medium at 30°C ( Figure 7A , middle and lower panels ) , which induces moderate filamentation , the homozygous sfl1 mutant displayed highly dense cell aggregates of a mixture of hyphae and long pseudohyphae ( Figure 7A , middle and lower panels ) , consistent with the function of SFL1 as a transcriptional repressor of filamentous growth . Interestingly , deletion of SFL2 or BRG1 in the sfl1 mutant strongly reduced filamentous growth as well as cell aggregation ( Figure 7A , middle and lower panels ) , with the sfl1 sfl2 double mutant cells growing as both yeast form and long to medium-size pseudohyphae and the sfl1 brg1 double mutants growing as both yeast form and short pseudohyphae ( Figure 7A , middle and lower panels ) . Single homozygous sfl2 and brg1 mutants showed phenotypes that were similar to those of the parental wild-type cells ( Figure 7A , middle and lower panels ) . We showed that Sfl2p directly upregulated UME6 and TEC1 expression ( Figures 3 , 5B and 6A ) , while specifically directly downregulating the expression of SFL1 ( Figures 3 , 5B and 6B ) , suggesting that SFL2 controls hyphal induction through at least UME6 , TEC1 and SFL1 . We tested the effect of overexpressing SFL2 on C . albicans morphogenesis in strains carrying the single homozygous deletions sfl1 , sfl2 , ume6 , tec1 , brg1 and efg1 ( Figure 7B ) . We and others previously showed that SFL2 overexpression in non-hypha-inducing conditions promotes hyphal growth [39] , [40] . We used the pNIMX system [41] to drive high levels of SFL2 expression in the above-mentioned strain backgrounds grown in rich medium ( Figure 7B ) . Overexpression of SFL2 in the wild-type strain strongly induced filamentation , with cells displaying long pseudohyphae ( Figure 7B , top panels ) . Interestingly , SFL2-driven filamentation was increased in the sfl1Δ/sfl1Δ mutant , as compared to that in the wild-type or the sfl2Δ/sfl2Δ strains ( Figure 7B , compare the zoomed-out regions in lower left corners ) . Most of the sfl1 mutant cells overexpressing SFL2 formed longer hyphae and pseudohyphae than those observed in the equivalent sfl2 mutants ( Figure 7B ) , suggesting that Sfl2p induces filamentous growth in part through repression of SFL1 expression . Conversely , filamentation was strongly reduced in the ume6Δ/ume6Δ strain , moderately reduced in either the tec1Δ/tec1Δ or brg1Δ/brg1Δ mutants and abolished in the efg1Δ/efg1Δ strain ( Figure 7B ) . The ume6 mutants overexpressing SFL2 formed significantly shorter pseudohyphae than those of the equivalent tec1 and brg1 mutants ( Figure 7B ) . Taken together , our results suggest that Sfl1p represses filamentation through at least direct negative regulation of SFL2 and BRG1 expression and indicate that Sfl2p regulates hyphal growth partly through UME6 , TEC1 and BRG1 and totally through EFG1 . Many observations support the hypothesis that Sfl1p and Sfl2p recognize different binding motifs . First , although sharing common transcriptional targets , Sfl1p and Sfl2p peak signals are distributed differently along many of their common target promoters ( Figure 2B , middle panel as an example ) . Second , Sfl2p binds specifically to the promoter of 75 targets ( Figure 2B , bottom panel as an example ) . Third , recent data by Song et al . suggested that Sfl1p and Sfl2p mediate their functional divergence through their HSF-type DNA binding domain [39] , suggesting divergent binding sites . We performed motif-enrichment analyses using DNA sequences encompassing ±250 bp around peak summits in Sfl1p ( Figure 8A ) and Sfl2p ( Figure 8B ) binding data . Two independent motif discovery algorithms , the RSA-tools ( RSAT ) peak-motifs ( http://rsat . ulb . ac . be/rsat/ , [55] ) and SCOPE ( genie . dartmouth . edu/scope/ , [56] ) were used ( See Materials and Methods for details ) . Strikingly , the highest scoring motifs in Sfl1p-enriched sequences included the Ndt80p ( 5′-ttACACAAA-3′ , mid-sporulation element , lowercase letters represent nucleotides with low-frequency occurrence ) and the Efg1p ( 5′-taTGCAta-3′ ) binding motifs [51] , [54] , [57] in addition to two high scoring motifs , 5′-TtCtaGaA-3′ and 5′-TCGAACCC-3′ , carrying GAA triplets that are characteristic of HSEs ( Figure 8A , shown are motifs found using the global overrepresentation of words relative to control sequences , significance index score ( i . e . −log10 E-value ) >10 for RSAT analyses and >25 for SCOPE analyses ) . Ndt80p is a transcription factor that controls the expression of genes involved in many cellular processes , including drug resistance , cell separation , morphogenesis and virulence through the recognition of mid-sporulation elements on the promoter of its targets [57] , [58] . This suggests the presence of functional interactions between Sfl1p , Efg1p and Ndt80p and proposes that Sfl1p binds to two different motifs or that an additional factor binds either 5′-TCGAACCC-3′ or 5′-TtCtaGaA-3′ . We searched the YeTFaSCo and the JASPAR databases for similarity with known transcription factor binding sites [59] , [60] . Interestingly , the 5′-TtCtaGaA-3′ sequence was strongly similar to the S . cerevisiae Hsf1p motif ( P = 3 . 856×10−04 , using YeTFaSco ) , while database searches did not identify any known motif that closely resembled the 5′-TCGAACCC-3′ sequence ( data not shown ) . On the other hand , we found 3 high-scoring motifs in Sfl2p-enriched sequences , including the Efg1p and Ndt80p binding motifs as well as the GAA-containing sequence , 5′-aaNAATAGAA-3′ ( where N represents any nucleotide; shown are motifs found using the position-analysis program , significance index score >5 ) ( Figure 8B ) . To confirm that the 5′-aaNAATAGAA-3′ motif was specific to Sfl2p , we performed motif discovery analyses using DNA sequences encompassing ±250 bp around peak summits of the regions specifically bound by Sfl2p and found the similar high-scoring motif 5′-aANAATAGAA-3′ ( Figure 8C ) . The 5′-aANAATAGAA-3′ motif shows moderate similarity with the S . cerevisiae Sfl1p and Mga1p motifs ( scores = 17 . 75 and 17 . 36 , respectively using the JASPAR database ) . All these identified motifs were distributed preferentially around the center of the sequences corresponding to peak locations ( Figures 8A , 8B and 8C ) , suggesting that Sfl1p , Sfl2p , Efg1p and Ndt80p binding sites were very close to each other . To determine if Efg1p and Ndt80p binding sites overlapped with the genome-wide occupancies of Sfl1p and Sfl2p , we compared Efg1p and Ndt80p binding profiles [51] , [57] to those of Sfl1p and Sfl2p ( Figure 8D ) . Ndt80p binding was resolved by Sellam et al . under yeast-form growth conditions at 30°C [57] , whereas Efg1p binding was analysed by Lassak et al . during both yeast-form growth ( 30°C ) and hyphal induction ( YP serum at 37°C ) [51] . Strikingly , a high proportion of Sfl1p and Sfl2p binding sites overlapped with those of Ndt80p ( Figure 8D ) , whereas Efg1p binding overlap was less frequent and depended on the morphological state of C . albicans , with rare or no overlap under hyphal induction and increased overlap under yeast-form growth ( Figure 8D ) . Roughly , 90% of Sfl1p and Sfl2p common targets were bound by both Ndt80p and Efg1p ( Figure 8D , upper panel as an example ) , whereas ∼10% ( 10 out of 113 common targets ) were bound by Ndt80p but not Efg1p . In at least two cases , Sfl1p and Sfl2p occupancy to common targets overlapped only with Efg1p binding: the promoter regions of SIS1 and PDE1 . On the other hand , ∼47% of Sfl2p specific targets were bound by both Ndt80p and Efg1p , whereas ∼42% overlapped only with Ndt80p binding ( Figure 8D , middle panel as an example ) . On rare occasions ( ∼11% ) , Sfl2p did not show significant overlap with the binding of any of the three regulators ( Figure 8D , bottom panel as an example ) . Taken together , our results indicate that Sfl1p and Sfl2p bind to DNA via divergent motifs and suggest the co-binding of transcription factors Efg1p and Ndt80p to many Sfl1p and Sfl2p target promoters , either concomitantly or successively , depending on growth conditions . Our bioinformatic analyses suggested the co-binding of Efg1p to many Sfl1p and Sfl2p target promoters . To test whether Sfl1p , Sfl2p and Efg1p concomitantly bind to common targets in vivo , strains individually expressing chromosomally TAP-tagged Sfl1p and Sfl2p ( strains SFL1-TAP and SFL2-TAP , Table 1 ) and HA-tagged Efg1p ( strain HLCEEFG1 , [18] , Table 1 ) under the control of their endogenous promoter were grown in SC medium at 30°C ( yeast form-promoting condition ) or in Lee's medium at 37°C ( filamentous form-promoting condition ) during 4 h before being subjected to ChIP-PCR analyses to detect differential binding of the three transcription factors to the promoter of selected Sfl1p and Sfl2p targets ( BRG1 , EFG1 , SFL2 , UME6 and TEC1 , Figure 9A , see Materials and Methods for details ) . All strains displayed similar hyphal growth phenotypes at 37°C in Lee's medium , whereas the yeast form growth phenotypes were similar for cells grown in SC medium at 30°C ( Figure S1A ) . Immunoblotting confirmed the expression of the different fusion proteins under the corresponding growth conditions ( Figure S1B ) . As expected , Sfl1p and Efg1p binding was detected at all tested promoters in SC medium at 30°C ( Figure 9A , compare lanes 1 and 7 to lanes 2 and 8 , respectively ) . Conversely , in Lee's medium at 37°C , Sfl1p and Efg1p binding was less efficient ( Figure 9A , Sfl1p binding , compare lanes 1 and 2 to lanes 4 and 5; Efg1p binding , compare lanes 7 and 8 to lanes 9 and 10 ) . Similarly , Sfl2p binding was detected at all tested promoters in Lee's medium at 37°C ( Figure 9A , compare lane 4 to lane 6 ) , whereas in SC medium at 30°C , Sfl2p binding was less efficient ( Figure 9A , compare lanes 4 and 6 to lanes 1 and 3 ) . To further explore the functional interaction between Sfl1p , Sfl2p and Efg1p , we sought to verify if the Efg1p protein could be co-immunoprecipitated with Sfl1p or Sfl2p in vivo . To this end , we generated strains co-expressing C-terminally TAP-tagged Sfl1p or Sfl2p and HA-tagged Efg1p ( AVL12-SFL1-TAP and AVL12-SFL2-TAP , respectively , Table 1 ) under the control of their chromosomal promoter together with control strains carrying individual Sfl1p-TAP , Sfl2p-TAP or Efg1p-HA fusions ( strains SFL1-TAP , SFL2-TAP and AVL12-pHIS , Table 1 , see Materials and Methods ) . Strains were grown during 4 h in SC medium at 30°C or in Lee's medium at 37°C , followed by crosslinking with formaldehyde to stabilize protein complexes and total extracts were incubated with IgG-coated beads for immunoprecipitation of the Sfl1p-TAP or Sfl2p-TAP proteins in the corresponding strain backgrounds . Immunoblotting with an anti-TAP antibody ( Figure 9B , IP , Anti-TAP panel ) allowed to detect the Sfl1p-TAP signal in beads incubated with extracts from strains carrying the SFL1-TAP allele irrespective of the growth conditions ( i . e . in both SC medium at 30°C and Lee's medium at 37°C ) ( Figure 9B , IP , Anti-TAP panel , lanes 2 , 4 , 7 and 9 ) . On the other hand , very low amounts of the Sfl2p-TAP protein fusion were detected in beads incubated with extracts from strains carrying the SFL2-TAP allele and grown in SC medium at 30°C ( Figure 9B , IP Anti-TAP panel , lanes 3 and 5 ) , however , the Sfl2p-TAP signal strongly increased in Lee's medium at 37°C ( Figure 9B , Anti-TAP panel , compare lanes 3 and 5 to lanes 8 and 10 ) . Interestingly , immunoblotting of the bound fractions with an anti-HA antibody ( Co-IP , Anti-HA panel ) allowed to detect Efg1p-HA co-immunoprecipitation with Sfl1p-TAP under both growth conditions: in SC medium at 30°C and in Lee's medium at 37°C ( Figure 9B , CoIP , Anti-HA panel , lanes 2 and 7 ) . Efg1p-HA co-immunoprecipitation with Sfl2p-TAP was barely detectable in SC medium at 30°C but was significantly enhanced in Lee's medium at 37°C , a condition that triggers increased expression of Sfl2p ( Figure 9B , CoIP , Anti-HA panel , compare lane 3 to lane 8 ) . As expected , Efg1p-HA was undetectable from beads incubated with strains individually expressing EFG1-HA , SFL1-TAP or SFL2-TAP ( Figure 9B , lanes 1 , 4 , 5 , 6 , 9 and 10 ) . Taken together , our results show that i ) the Efg1p protein binds to many Sfl1p and Sfl2p targets , in vivo and ii ) Both Sfl1p and Sfl2p proteins physically associate with Efg1p , in vivo .
The ChIP-Seq and transcriptomics technologies are powerful in vivo approaches that , when combined , allow to provide mechanistic insights into the function of transcriptional regulators . When associated with both genetic and physical interaction analyses , the overall generated data are cross-validated and provide a comprehensive view of the regulatory interactions within transcriptional networks . They also shed more light into the epistatic relationships to explain the phenotypes associated with transcription factor function . In the present report , we used such approaches to decipher the regulatory network of two HSF-type transcription factors , Sfl1p and Sfl2p , both required for C . albicans virulence but with antagonistic functions in regulating C . albicans morphogenesis . One limitation of our ChIP-Seq design was the use of ectopic promoter-driven expression of the SFL1-HA3 and SFL2-HA3 alleles ( Figure 1 ) . This may drive non physiological expression levels and some of the transcriptional changes and promoter occupancies may be altered from the situation where the genes are expressed from their endogenous promoters . Nevertheless , phenotypic analyses suggested that at least PMET3-driven expression of SFL2-HA3 imparts filamentous growth in a manner similar to the wild-type SC5314 strain ( Figure 1C ) . Furthermore , we generated strains expressing TAP-tagged SFL1 and SFL2 from their endogenous promoter and ChIP experiments using these strains confirmed some of our data that used the PMET3 expression system ( Figure 9A ) . Our data allow to propose a model of Sfl1p and Sfl2p transcriptional network ( Figure 10 , for simplicity only binding associated with transcriptional modulation is shown ) as well as a mechanism whereby Sfl1p and Sfl2p antagonistically regulate the yeast-to-hyphae transition ( see below ) . Sfl2p , which responds to temperature increase , and Sfl1p bind to the promoter of common target genes ( blue boxes in Figure 10 ) belonging to at least 3 functional groups involved in morphogenesis: transcriptional repressors of hyphal growth ( SSN6 , NRG1 , RFG1 , others ) , transcriptional activators of hyphal growth ( BRG1 , UME6 , TEC1 , others ) and yeast-form associated genes ( RME1 , RHD1 , YWP1 , others ) . While Sfl1p exerts direct negative and positive regulation on the expression of activators ( BRG1 , UME6 , TEC1 ) and repressors ( SSN6 , NRG1 ) of hyphal growth , respectively , Sfl2p directly upregulates and downregulates the expression of positive ( UME6 , TEC1 ) and negative ( RFG1 , NRG1 ) regulators of hyphal growth , respectively ( Figure 10 ) . Additionally , Sfl1p directly upregulates the expression of yeast-form associated genes ( RME1 , RHD1 and YWP1 ) whereas Sfl2p directly downregulates their expression ( Figure 10 ) . Moreover , Sfl1p and Sfl2p directly negatively regulate the expression of each other ( Figure 10 ) . As stated above , this model is consistent with the genetic interaction analyses performed between SFL1 ( genetically interacts with at least BRG1 and SFL2 ) , SFL2 ( genetically interacts with at least UME6 , TEC1 and BRG1 ) and their target genes ( Figure 7 ) . Importantly , on the other hand Sfl2p exclusively binds to the promoter of specific target genes that belong to at least 2 functional groups involved in morphogenesis: HSGs ( ALS3 , HGC1 , HWP1 , HYR1 , ECE1 , SAP4 , IHD1 , FAV2 , RBT4 ) and yeast-form specific genes ( PIR1 , RHD3 ) ( Figure 10 ) . We propose that binding of Sfl1p and Sfl2p to a high proportion of their transcriptional targets occurs with additional binding of transcription factors Ndt80p and/or Efg1p , depending on growth conditions ( Figures 8 , 9 and 10 ) , presumably through direct or indirect physical interaction ( Figures 8 and 9 , see below ) . One could speculate that the requirement of a functional EFG1 gene for Sfl1p and Sfl2p abilities to regulate morphogenesis under specific growth conditions ( Figure 7 and [39] ) could be explained by the need for Efg1p co-binding and/or physical interaction , as suggested by our study ( Figures 7 , 8 and 9 ) . Indeed , we show here that Efg1p co-immunoprecipitates , in vivo , with Sfl1p and Sfl2p and binds to the promoter of many Sfl1p and Sfl2p target genes ( Figure 9 ) . On the other hand , our finding that Sfl2p binds exclusively to specific targets , including a high proportion of HSGs ( Figure 6 ) , provides additional insight into SFL2 function . This might explain , for instance , why SFL2 was able to bypass the need of EFG1 and FLO8 to induce hyphal growth in embedded conditions at 37°C [39] . We are currently testing whether Sfl1p and Sfl2p binding to their targets requires the presence of functional EFG1 or NDT80 genes . Overall , we propose that the execution of these single ( including SFL1-SFL2 cross-factor negative control ) and multiple input motifs in Sfl1p or Sfl2p transcriptional network dictates the commitment of the C . albicans cells to form hyphae or yeast-form cells . This model is consistent with Sfl1p and Sfl2p acting as “switch on/off” proteins , with Sfl1p directly turning off the expression of positive regulators of hyphal growth while turning on the expression of both yeast-form associated genes and genes encoding repressors of hyphal development , whereas Sfl2p directly turns on the expression of HSGs and positive regulators of hyphal growth while turning off the expression of yeast-form associated genes as well as negative regulators of hyphal development ( Figure 10 ) . The mechanisms whereby HSF-type transcription factors activate transcription involve homotrimerization , post-translational modifications ( e . g . phosphorylation , others ) as well as interaction with multiple protein partners , followed by recruitment of the co-activating mediator complex and initiation of the transcriptional process [61] . This mechanism may include or not nuclear translocation , as many HSFs were shown to reside in the nucleus under both activating and non-activating conditions or to be imported to the nucleus following activation [61] . It was shown that Sfl1p is constitutively localized to the nucleus under both yeast- and hyphae-promoting conditions and irrespective of temperature levels [37] , [38] , whereas an Sfl2p-GFP fusion was undetectable at 25°C but displayed nuclear localization at 37°C [39] . Moreover , SFL2 RNA levels were undetectable by Northern blotting at either 25°C or 30°C , but were greatly enhanced upon temperature increase [39] and this correlated with Sfl2p protein level variations [39] . Indeed , we show here that in SC medium at 30°C , Sfl2p protein levels are low , but are significantly enhanced upon temperature increase to 37°C in Lee's medium ( Figure S1B ) . Moreover , we show that Sfl2p binding is more stable at 37°C in Lee's medium as compared to 30°C in SC medium , and vice versa for Sfl1p ( Figure 9A ) . Based on these observations , we propose the following model of Sfl1p/Sfl2p activation: Sfl1p binds to its transcriptional targets to maintain the yeast form growth at low temperature by directly modulating the expression of genes involved in morphogenesis ( Figure 10 ) . A temperature increase to 37°C leads to an increase in both Sfl2p expression and binding to the promoter of Sfl1p targets in addition to specific targets ( including HSGs ) and induction of the hyphal development program ( Figure 10 ) . As we show here that Sfl1p and Sfl2p act as both activators and repressors of gene expression ( Figures 6 and 10 ) , it is likely that they alternatively recruit ( directly or indirectly ) co-repressors ( e . g . Tup1p-Ssn6p ) and co-activators ( e . g . mediator-Swi/Snf complex ) at different binding sites to regulate morphogenesis . Our observation that Sfl2p binds to its own promoter , but not Sfl1p ( Figures 3 , 6Aand 10 ) is consistent with this model as SFL2 may undergo auto-induction which would lead to a rapid , amplified and sustained expression of SFL2 , allowing an efficient response to temperature increase . On the other hand , SFL1 expression , protein levels and nuclear localization remain constant under various conditions [38] , which may dispense the need for autoregulation . The SFL1-SFL2 cross-factor negative control is also consistent with this model . Under low temperature conditions , Sfl1p directly turns off SFL2 expression to prevent activation of hyphal growth . Upon a temperature increase , SFL2 expression is enhanced and Sfl2p binds to the SFL1 promoter to turn off SFL1 expression . This allows to relieve Sfl1p-mediated repression , thus contributing to activation of the hyphal development program . Our motif discovery analyses suggested that Ndt80p co-binds together with Efg1p to the promoter of Sfl1p and Sfl2p targets ( Figure 8 ) . We also strikingly found that a high proportion of Sfl1p and Sfl2p binding sites overlapped with those of Ndt80p and/or Efg1p ( Figure 8 ) . However , since the Ndt80p ChIP-on-chip was performed on yeast-form grown cells at 30°C [57] , one cannot exclude the possibility that Ndt80p binding is altered/lost upon hyphal induction , as is obviously the case for Efg1p ( [51] and Figures 8D and 9A ) . Ndt80p occupies the promoter region of roughly a quarter of total C . albicans genes under yeast-form growth conditions , suggesting wide functions for Ndt80p [57] . Indeed , it was shown that Ndt80p regulates different processes including drug resistance , cell separation , hyphal differentiation , biofilm formation and virulence [54] , [57] , [58] . Importantly , the C . albicans ndt80Δ/ndt80Δ mutant is unable to form true hyphae under different filamentation-inducing conditions and , in the presence of serum at 37°C , it fails to activate the expression of HSGs , including HWP1 , ECE1 , RBT4 , ALS3 , HYR1 and SAP4 [58] , all directly regulated by Sfl2p ( Figure 6 ) , as well as the transcription factor-encoding genes TEC1 and UME6 which are both directly modulated by Sfl1p and Sfl2p ( Figure 6 ) . Additionally , under the same growth conditions , the homozygous ndt80 mutant was unable to downregulate the yeast form-associated genes YWP1 , RHD3 , RHD1 and the transcriptional repressor-encoding gene NRG1 [58] , which are also direct targets of Sfl1p or Sfl2p ( Figure 6 ) . These observations , together with our findings that i ) Ndt80p binding motif was enriched among Sfl1p and Sfl2p bound sequences and that ii ) a significant proportion of its genome-wide binding profile overlapped with Sfl1p and Sfl2p binding , suggest that Sfl1p , Sfl2p and Ndt80p cooperatively regulate C . albicans morphogenesis in response to temperature variation . Whether Sfl1p and Sfl2p regulate this process through physical interaction with Ndt80p and the associated sequence of molecular events occurring during the yeast-to-hyphal switch await further characterization . On the other hand , we found that Efg1p binding also overlapped with that of Sfl1p and Sfl2p , at a lesser extent , though , as compared to Ndt80p binding ( Figure 8 ) . It is intriguing that Efg1p binding undergoes alteration following the induction of hyphal development ( [51] and Figures 8D and 9A ) . Our examination of Efg1p binding data by Lassak et al . [51] together with our ChIP experiments ( Figure 9A ) suggest that Efg1p binding to many targets is decreased/altered upon hyphal induction . We show here that during yeast-form growth , at low temperature , Efg1p co-immunoprecipitates with Sfl1p but not with Sfl2p , presumably due to the low levels of Sfl2p at low temperature ( Figure 9B ) . One could speculate that , at low temperature , Sfl1p associates directly or indirectly with Efg1p on the promoter of its targets to repress hyphal development . Following a temperature increase , both Sfl2p levels and Sfl2p DNA binding are enhanced ( Figures S1 and 9A ) , which in turn activates the hyphal development program . Although Efg1p binding is altered upon hyphal induction , Efg1p co-immunoprecipitated with Sfl2p ( Figure 9B ) at 37°C in Lee's medium , which may explain Sfl2p dependency on EFG1 to regulate morphogenesis under certain conditions . Nobile et al . elegantly showed that an intricate transcriptional network involving Ndt80p , Efg1p , Brg1p , Bcr1p , Rob1p and Tec1p controls biofilm development in C . albicans [54] . Interestingly , with the exception of BCR1 , all genes encoding these regulators are direct targets of Sfl1p or Sfl2p ( Figure 6 and [54] ) . It is tempting to speculate that Sfl1p and Sfl2p may convey temperature regulation to the transcriptional network controlling biofilm formation . C . albicans adaptation to temperature variation is one of the major critical traits of its ability to cause disease or to act as a commensal of warm-blooded species , as a temperature increase triggers hyphal development [2] . To date , three temperature-responsive transcription factors have been shown to play a role in C . albicans morphogenesis , Hsf1p [62] , [63] , Sfl2p [39] , [40] and Hms1p [49] . Importantly , all three transcription factors are required for full virulence in different host/tissue models [39] , [40] , [49] , [63] , reinforcing the link between temperature adaptation and pathogenesis in C . albicans . The HMS1 gene , encoding a basic helix-loop-helix ( bHLH ) transcription factor , has been recently isolated in a screen aimed at identifying transcription factors whose function is required for the HSP90- or high temperature-mediated filamentous growth [49] . Hms1p acts downstream of the Pho85p-Pcl1p cyclin-dependent kinase pathway but its function was still dependent upon cAMP-PKA signalling [49] . Interestingly , both Sfl1p and Sfl2p bind to the promoter of the HMS1 gene , while Sfl2p downregulates its expression ( Figure 6A ) , suggesting that activation of Sfl2p turns off the HSP90-dependent filamentation response ( at least under the conditions used in the present study ) . Similar to Sfl2p , Hsf1p is an HSF-type transcription factor that induces transcription following a temperature increase , but , unlike SFL1 and SFL2 , HSF1 is essential for viability [62] . Hsf1p is required for the expression of essential chaperones , including HSP104 , HSP90 , HSP70 as well as other classical heat-shock protein ( HSP ) -encoding genes such as HSP60 , HSP78 , others [62] . Although carrying HSF-type domains in their primary protein sequences and sharing relatively high sequence similarity levels with Hsf1p , speculating a role in the transcriptional regulation of HSP ( or HSP-related ) genes , the Sfl1p and Sfl2p binding targets did not show any significant enrichment of functional categories pertaining to the heat-shock response pathway ( e . g . protein folding/refolding ) , including HSPs and chaperones ( Figure 2C ) . This may have important evolutionary implications as it might reflect specific needs of C . albicans to efficiently act as an opportunistic yeast of warm-blooded animals through converting temperature-sensing inputs into a morphogenesis programming output using HSF-type regulators like Sfl1p and Sfl2p . Nevertheless , we detected Sfl1p and Sfl2p binding at the promoter of the HSP104 , HSP70 and SIS1 genes ( binding intensity below algorithm threshold used for HSP70 ) , suggesting that a reminiscent classical heat-shock response may have been retained in Sfl1p and Sfl2p . It is intriguing that one of the two potential binding motifs of Sfl1p ( Figure 8A ) , 5′-TtCtaGaA-3′ , is strikingly similar to the S . cerevisiae Hsf1p motif [64] , [65] , in line with the hypothesis that transcriptional rewiring affected the regulation of the heat shock response and temperature adaptation between S . cerevisiae and C . albicans . It is worth noting that the predicted protein sequences of Sfl1p and Sfl2p are highly similar to those of S . cerevisiae Sfl1p and Mga1p . The MGA1 gene has been initially isolated as a multicopy suppressor of both the snf2Δ ( component of the SWI/SNF remodelling complex , also known as gam1 ) [66] and the mep1Δ/mep1Δ mep2Δ/mep2Δ ( encoding ammonium permeases ) filamentous defect [67] mutations in S . cerevisiae . Interestingly , Mga1p was shown to act as a master regulator of S . cerevisiae pseudohyphal development through direct transcriptional control of key genes involved in morphogenesis [68] . Many intriguing functional similarities exist between Sfl2p and S . cerevisiae Mga1p , although either SFL1 or SFL2 could complement an sfl1Δ mutation and SFL2 could not complement the pseudohyphal growth defect of an mga1Δ mutant [39] . First , both proteins recognize similar DNA binding motifs ( 5′-AtAGAACA-3′ for Mga1p [33] and 5′-ANATAGAA-3′ for Sfl2p ( Figure 8 ) ) . Second , both transcription factors bind to the promoter of orthologous genes ( ScPHD1 and ScSOK2/CaEFG1 , HMS1 , ScGAT2/CaBRG1 , MSB2 , ACH1 , ScENA1/CaENA21 , GCN4 , CUP9 , TPO4 , ScSCW4/CaMP65 , others; binding to some genes is below peak-finding algorithm threshold ) . Third , the regulatory networks to which they belong are intriguingly similar: Mga1p establishes cross talks with major regulators of S . cerevisiae pseudohyphal growth including Phd1p , Sok2p ( Efg1p orthologs ) , Flo8p and Tec1p , as in the case of Sfl2p ( Figure 6 ) [39] , [68] . Fourth , overexpression of MGA1 and SFL2 is sufficient to induce morphogenesis in the respective species under conditions that do not promote filamentation [39] , [68] . Fifth , Sfl2p requires EFG1 and FLO8 to induce filamentation under specific conditions ( Figure 7B and [39] ) and we show here that Efg1p co-immunoprecipitates with Sfl2p ( Figure 9B ) . Similarly , Mga1p requires a functional FLO8 gene for its ability to bind DNA and Mga1p and Flo8p interact with each other [68] . We suggest that transcriptional rewiring may have affected the functions of Sfl2p and Mga1p in their respective species: In diploid S . cerevisiae cells , Mga1p responds to nitrogen limitation to turn on pseudohyphal growth , whereas in C . albicans Sfl2p responds to temperature increase to induce hyphal development .
The C . albicans strains used in this study are listed in Table 1 . Depending on experimental conditions , C . albicans strains were grown in YPD ( 1% yeast extract , 2% peptone , and 1% dextrose ) , YP ( 1% yeast extract , 2% peptone ) supplemented with 10% Fetal Bovine Serum ( FBS ) , SD ( synthetic dextrose , 0 . 67% yeast nitrogen base ( YNB; Difco ) with 2% glucose ) [69] supplemented if necessary with arginine , histidine or uridine ( 20 mg/l each and 2% agar for growth on solid medium ) , SC ( synthetic complete ) or Lee's medium supplemented or not with methionine [70] . Expression from the tetracycline-inducible promoter ( PTET ) was achieved through addition of 3 µg/ml anhydrotetracycline ( ATc - Fisher Bioblock Scientific ) in YPD at 30°C [41] . ATc-containing cultures were maintained in the dark as ATc is light sensitive . Escherichia coli strains TOP10 ( Invitrogen ) or DH5α were used for DNA cloning and maintenance of the plasmid constructs . All C . albicans transformation experiments used the lithium-acetate transformation protocol of Walther and Wendland [71] and selection of transformants for uridine or histidine prototrophy ( when using the URA3 or the HIS1 markers , respectively ) or Nourseothricine resistance ( when using the SAT1 marker ) [72] . Plasmid pCaMPY-3xHA and the SGY243 strains expressing the CAP1-HA3 allele or carrying the empty vector ( pCaEXP ) were kindly provided by Dr Martine Raymond ( Université de Montréal , Canada ) . Strains AVL12 and HLCEEFG1 ( expressing EFG1-HA under the control of the endogenous promoter ) were the kind gifts of Dr Joachim Ernst ( Heinrich-Heine-Universität , Dusseldorf , Germany ) . We first attempted to generate epitope ( HA3 , triple hemagglutinin ) -tagged strains expressing Sfl1-HA3 or Sfl2-HA3 under the control of their endogenous promoter at their chromosomal location . SFL1- or SFL2-tagging cassettes were PCR-amplified from plasmid pCaMPY-3×HA [73] using primers SFL1-HA-FWD ( forward , Table S9 in Text S1 , the lowercase sequence corresponds to positions +2316 to +2415 of the SFL1 ORF ) and SFL1-HA-REV ( reverse , Table S9 in Text S1 , the lowercase sequence corresponds to positions +2419 to +2518 of the SFL1 ORF ) or primers SFL2-HA-FWD ( forward , Table S9 in Text S1 , the lowercase sequence corresponds to positions +2043 to +2142 of the SFL2 ORF ) and SFL2-HA-REV ( reverse , Table S9 in Text S1 , the lowercase sequence corresponds to positions +2146 to +2245 of the SFL2 ORF ) , which anneal specifically to the in-frame pCaMPY-3×HA vector sequences PET-up and PET-down ( respective uppercase sequences in Table S9 in Text S1 ) , as described previously [73] . The resulting fragments ( 1 , 853 bp ) , containing the C . albicans URA3 marker flanked by direct repeats of the HA3-encoding sequences and 100 bp of sequences homologous to the 3′ end of the SFL1 or SFL2 genes , were used to respectively transform ura3-deficient sfl1Δ/SFL1 and sfl2Δ/SFL2 heterozygous mutants , yielding strains CEC3075 and CEC3076 , respectively ( Table 1 ) . Expression of the Sfl1p-HA3 and Sfl2p-HA3 fusions in strains CEC3075 and CEC3076 was not detectable by Western blot analyses , suggesting that integration of the tagging cassette at the 3′ untranslated regions of SFL1 and SFL2 had a knockdown effect . Despite many attempts , excision of the URA3 marker through intramolecular recombination between the HA3 sequences was not successful . We rather observed 100% loss of the entire tagging cassette at the SFL1 and SFL2 loci . We therefore used the pCaEXP system to drive expression of the tagged SFL1 and SFL2 alleles at the RPS1 locus [42] . The SFL1-HA3 or SFL2-HA3 fusions were PCR amplified from CEC3075 or CEC3076 genomic DNA , respectively , using primers SFL1-HA-CaEXP-FWD ( forward , Table S9 in Text S1 , introduces a BglII site [underlined] ) or SFL2-HA-CaEXP-FWD ( forward , Table S9 in Text S1 , introduces a BglII site [underlined] ) , respectively , and primer HA-CaEXP-REV ( reverse , Table S9 in Text S1 , introduces sequentially a BglII site [underlined] and a TAA stop codon [in red lowercase letters] ) . The resulting fragments ( SFL1-HA3 , ∼2 , 600 bp; SFL2-HA3 , ∼2 , 330 bp ) were digested with BglII and cloned into the compatible BamHI site of plasmid pCaEXP , generating plasmids pCaEXP-SFL1-HA3 and pCaEXP-SFL2-HA3 . Plasmids pCaEXP ( empty vector , control ) , pCaEXP-SFL1-HA3 and pCaEXP-SFL2-HA3 were digested with StuI for integration at the RSP1 locus [42] and the resulting fragments were used to transform strains CEC1910 and CEC1503 ( Table 1 ) , respectively , to generate strains sfl1-CaEXP , sfl1-CaEXP-SFL1-HA3 , sfl2-CaEXP and sfl2-CaEXP-SFL2-HA3 ( Table 1 ) . Construction of C . albicans knock-out mutants ( Table 1 ) used PCR-generated ARG4 , HIS1 , URA3 and SAT1 disruption cassettes flanked by 100 base pairs of target homology region ( primer sequences are listed in Table S9 in Text S1 ) as described by Gola et al . [74] and Schaub et al . [75] . Independent transformants were produced and the gene replacements were verified by PCR on whole yeast cells as described previously [74] , [75] . If necessary , transformants were converted to uracil prototrophy using StuI-linearized CIp10 [76] . Mutant strains carrying the pCIp-PTET-SFL2 [41] plasmid ( Table 1 ) were first transformed with the pNIMX construct as described in Chauvel et al . [41] . Construction of chromosomally TAP-tagged SFL1 and SFL2 alleles ( Table 1 ) used PCR-generated tagging cassettes from plasmid pFA-TAP-HIS , a derivative of the pFA-GFP-tagging plasmid series [74] ( primers are listed in Table S9 in Text S1 , oligos # 50-53 ) followed by targeted homologous recombination at the 3′ untranslated regions of SFL1 and SFL2 to generate strains expressing C-terminally tagged Sfl1p ( strains SFL1-TAP and AVL12-SFL1-TAP , Table 1 ) and Sfl2p ( strains SFL2-TAP and AVL12-SFL2-TAP , Table 1 ) proteins . Total protein extracts were prepared from 24 OD600 units of strains expressing ( sfl1-CaEXP-SFL1-HA , sfl2-CaEXP-SFL2-HA ) or not ( empty vector; sfl1-CaEXP , sfl2-CaEXP ) the SFL1-HA3 or SFL2-HA3 alleles ( Table 1 ) grown overnight in SD medium ( PMET3-inducing conditions ) . Cultured cells were centrifuged at 3 , 500 rpm during 5 min at room temperature and the pellets were resuspended in 150 µl of ice-cold TE buffer ( 10 mM Tris , [pH 7 . 5] , 1 . 5 mM EDTA ) supplemented with a protease inhibitor cocktail ( Roche ) and 1 . 5 mM phenylmethylsulfonyl fluoride ( PMSF ) then transferred to 1 . 5-ml tubes . The equivalent of 100 µl ice-cold glass beads was added to each tube and the suspensions were vortexed 5 times during 1 minute with 1-min incubations on ice in between . The extracts were clarified by centrifugation at 5 , 000 rpm during 1 min , boiled for 1 min and separated ( 25 µl ) by electrophoresis on a sodium dodecyl sulfate-8% polyacrylamide gel . Proteins were electrophoretically transferred to nitrocellulose membranes . The membranes were incubated with a mouse anti-HA monoclonal antibody ( 12CA5; Roche ) for 1 h at a dilution of 1∶1 , 000 , followed by incubation with a horseradish peroxidase-conjugated secondary antibody ( Sigma ) during 30 min , washed , and developed with enhanced chemiluminescent detection reagents ( ECL kit , GE Healthcare ) . Cells were observed with a Leica DM RXA microscope ( Leica Microsystems ) . Images were captured with a Hamamatsu ORCA II-ER cooled CCD camera , using the Openlab software version 3 . 5 . 1 ( Improvision Inc . ) . Two independent cultures of strains sfl1-CaEXP or sfl2-CaEXP ( untagged; control strains ) and sfl1-CaEXP-SFL1-HA3 or sfl2-CaEXP-SFL2-HA3 ( tagged strains ) ( Table 1 ) were grown overnight in 2 ml YPD at 30°C , diluted to an OD600 of 0 . 3 in Lee's medium deprived of methionine and cysteine ( to induce PMET3 ) and grown during 4 hours at 37°C ( hyphae-inducing conditions ) . The subsequent steps of DNA cross-linking , DNA shearing and chromatin immunoprecipitation ( ChIP ) were conducted as described in Liu et al . [73] , with some modifications . Briefly , cultures were treated with 1% formaldehyde ( cross-linking ) and snap-frozen in liquid nitrogen . Total cell extracts were prepared by bead beating using a FastPrep-24 instrument ( MP Biomedicals ) with 6 runs during 1 minute each at 6 . 0 m/sec and 1 minute on ice in between ( these settings led to efficient breakage of hyphal cells ) . Preparation of soluble chromatin fragments was performed by sonicating the extracts 6 times during 20 sec at power 8 ( knob position ) for an output signal amplitude of 15 ( Microns , Peak to Peak ) using a probe sonicator ( MSE ) , yielding ∼200-bp DNA fragments on average . The extracts were then incubated at 4°C overnight with a mouse monoclonal anti-HA antibody ( Santa Cruz Biotech ) coupled to magnetic beads ( pan-mouse immunoglobulin G Dynabeads; Dynal Biotech , Brown Deer , WI ) . The concentration of the purified immunoprecipitated DNA was ranging between 0 . 2 ng/µl and 1 . 5 ng/µl in 50 µl TE ( 10 mM Tris [pH 8 . 0] , 1 mM EDTA ) . Library construction ( 10 ng of the immunoprecipitated DNA were used , adaptor-DNA fragments ranging from 150 to 350 bp ) was performed using the TruSeq DNA sample preparation kit as recommended by the manufacturer ( Illumina ) , followed by quality control analyses using a Bioanalyzer 2100 instrument ( Agilent Technologies ) . DNA library samples were indexed and pools of the Sfl1p ( 4 samples , both tagged and control ) or Sfl2p ( 4 samples , both tagged and control ) ChIP samples were loaded onto two lanes of an Illumina HiSeq2000 sequencer flow cell for single-read ( 51 base pairs per read ) high-throughput sequencing . The resulting 51-nucleotide sequence reads ( FASTQ files ) were imported into the Galaxy NGS data analysis software ( https://main . g2 . bx . psu . edu/ ) and the tools implemented in Galaxy were used for further processing via workflows [77] , [78] . Quality control analyses of the FASTQ files were performed using FastQC ( version 0 . 10 . 0 , Babraham Institute ) and adaptor-contaminated sequences were trimmed . The reads were then mapped to the C . albicans assembly 21 genome using the Bowtie algorithm [79] and the files of mapped reads ( BAM files ) for the ChIP sample ( 2 biological replicates from samples sfl1-CaEXP-SFL1-HA3 or sfl2-CaEXP-SFL2-HA3 ) and from the control ( 2 biological replicates from samples sfl1-CaEXP or sfl2-CaEXP ) were processed using the command line version 1 . 4Orc2 of the Model-Based Analysis for ChIP-Seq ( MACS ) peak-finding algorithm [46] for peak finding with the following parameters: bandwidth = 250; mfold = 10 , 30; shiftsize = 100; P-value cutoff for Sfl1p peaks = 1e-14 and P-value cutoff for Sfl2p peaks = 1e-100 . Replicates 1 and 2 from the two independently performed ChIP-Seq experiments were processed separately . Overlapping peak intervals ( intersection ) from replicates 1 and 2 of Sfl1p or Sfl2p binding data were generated using the Galaxy tool Intercept version 1 . 0 . 0 ( https://main . g2 . bx . psu . edu/ ) . The complete Sfl1p and Sfl2p binding and expression datasets are provided in Tables S1–S8 in Text S1 . The command line version of the PeakAnnotator ( v 1 . 4 ) sub-package from the PeakAnalyzer suite of algorithms [80] was used to annotate the Sfl1p and Sfl2p binding peaks in Tables S1 , S2 , S4 and S5 in Text S1 . The association of peaks to target genes was also conducted by human eye ( Tables S3 and S6 in Text S1 ) , based on the location of ORFs relative to binding peaks . We provide wiggle tracks with tag counts for every 10 bp segment ( See Materials and Methods section entitled “Data accession numbers” below ) . Visualization of the ChIP-Seq results was conducted using the Integrated Genomics Viewer software [44] , [45] . Thirty cycles of PCR with 15 seconds at 95°C , 15 seconds at 50°C and 40 seconds at 70°C were performed on independently generated ChIP samples ( Figures 3 and 9A ) in a 50-µl reaction volume with 1 µl ( 5% ) of immunoprecipitated material . Primers were designed to assay binding enrichment approximately around ChIP-Seq peak summits ( primer sequences are provided in Table S9 in Text S1 ) . The URA3 and YAK1 ORFs were used as negative controls . Strains sfl1-CaEXP or sfl2-CaEXP ( control strains , for subsequent Cy3 labeling ) and sfl1-CaEXP-SFL1-HA3 or sfl2-CaEXP-SFL2-HA3 ( test strain , for subsequent Cy5-labeling ) ( Table 1 ) were grown overnight in 2 ml YPD at 30°C . The next day , an aliquot of the overnight culture was used to inoculate 50 ml of Lee's medium deprived of methionine and cysteine to a starting OD600 of 0 . 3 . This culture was grown for 4 hours at 37°C , cells were washed with diethyl pyrocarbonate ( DEPC ) -treated water , collected by centrifugation and pellets were immediately frozen and stored at −80°C until RNA isolation . Three independently obtained sets of cell cultures were used . RNA was isolated from frozen cell pellets using the hot-phenol method [81] . Briefly , cells were resuspended in 375 µl TES buffer ( 10 mM Tris [pH 7 . 5] , 10 mM EDTA , 0 . 5% SDS ) at room temperature , after which 375 µl acid Phenol∶Chloroform ( 5∶1 , Amresco , Solon , OH ) were added . Samples were then incubated for 1 hour at 65°C with vigorous vortexing during 20 sec each 10 min and subjected to centrifugation for 20 min at 14 , 000 rpm . The supernatants were transferred to new tubes containing 750 µl acid Phenol∶Chloroform ( 5∶1 ) , mixed , and subjected to centrifugation at 14 , 000 rpm for 10 min . The aqueous phase was transferred to new tubes containing 750 µl Chloroform∶Isoamyl alcohol ( 24∶1 , Interchim , Montluçon , France ) , mixed and centrifuged at 14 , 000 rpm during 10 min . RNA was precipitated from the resulting aqueous layer by mixing that portion in new tubes with 1 ml 99% ethanol ( pre-cooled at −20°C ) and 37 µl of 3 M sodium acetate [pH 5 . 0] and subjecting the mixture to centrifugation at 14 , 000 rpm for 40 min at 4°C . The supernatants were removed , the pellet was resuspended in 500 µl 70% ethanol , and the RNA was collected by centrifugation at 14 , 000 rpm for 20 min at 4°C . The supernatants were again removed , and the RNA was resuspended in 150 to 300 µl DEPC-treated water . The RNA was stored at −80°C until needed . Prior to first-strand cDNA synthesis , the purity and concentration of RNA samples were determined from A260/A280 readings ( NanoVue Plus , GE Healthcare ) and RNA integrity was determined by a Bioanalyzer 2100 instrument ( Agilent Technologies ) per the manufacturer's instructions ( RNA concentration was ranging between 7 . 92 and 10 . 48 µg/µl ) . First-strand cDNA was synthesized from 20 µg total RNA , using the Superscript III indirect cDNA labeling system ( Invitrogen ) with the following minor modifications to the manufacturers' instructions . Briefly , the Qiagen PCR Purification kit was used to remove unincorporated aminoallyl-dUTP and free amines with substitution of the Qiagen-supplied buffers with phosphate wash ( 5 mM Phosphate buffer [K2HPO4/KH2PO4O4] [pH 8 . 0] , 80% ethanol ) and elution ( 4 mM Phosphate buffer [K2HPO4/KH2PO4O4] [pH 8 . 5] ) buffers . The purified first-strand cDNAs were subsequently labelled with the monoreactive Cy dye N-hydroxysuccinimide esters Cy3 ( control , cDNA from strains sfl1-CaEXP or sfl2-CaEXP ) and Cy5 ( cDNA from strains sfl1-CaEXP-SFL1-HA3 or sfl2-CaEXP-SFL2-HA3 ) ( GE Healthcare ) and the uncoupled dye was removed using the standard Qiagen PCR purification kit protocol . The Cy3- and Cy5-labeled cDNA lyophilized pellets were resuspended in 10 µl of DNase-free water then 2 . 5 µl and 12 . 5 µl of 10X blocking agent and 2X hybridization buffer ( Agilent Technologies ) , respectively , were added . The resulting samples were mixed , incubated at 95°C during 3 min and snap cooled on ice during 1 min then hybridized to a Candida albicans expression array ( Agilent Technologies ) designed such that two nonoverlapping probe sets are targeting each of 6 , 105 C . albicans ORFs for a total of 15 , 744 probes , thereby allowing two independent measurements of the mRNA level for a given gene ( The EMBL-European Bioinformatics Institute ArrayExpress platform accession number: A-MEXP-2142 , http://www . ebi . ac . uk/arrayexpress/arrays/A-MEXP-2142 ) . Images of Cy5 and Cy3 fluorescence intensities were generated by scanning the expression arrays using an Axon Autoloader 4200AL scanner ( Molecular Devices , Downington , PA ) . Images were subsequently analyzed with the GenePix Pro 6 . 1 . 0 . 2 software ( Molecular Devices , Downington , PA ) . GenePix Results ( GPR ) files were imported into the Arraypipe 2 . 0 [82] or the GeneSpring ( Agilent Technologies ) softwares . Following spot filtering and bad spot flagging , global signal intensities were normalized using Loess normalization and replicate slides ( n = 3 ) were combined and the P-values calculated using a standard Student's t-test . Total RNA was prepared from strains CEC2001 ( sfl1Δ/sfl1Δ ) and CEC1997 ( sfl1Δ/sfl1Δ PPCK1-SFL1-TAP ) or CEC1535 ( sfl2Δ/sfl2Δ ) and CEC1509 ( sfl2Δ/sfl2Δ PPCK1-SFL2-TAP ) ( Table 1 ) during a kinetics experiment ( 0 h , 2 h and 4 h ) in YNB plus 2% casaminoacids ( PPCK1-inducing conditions ) . Cells from 100 mL cultures were mechanically disrupted with glass beads using a Fastprep ( MP Biomedicals ) and total RNA was extracted using RNAeasy ( QIAGEN ) according to the manufacturer's instructions . The quality and quantity of the isolated RNA were determined using an Agilent 2100 Bioanalyzer . Before cDNA synthesis , total RNA samples were DNase-treated using the Turbo DNA-free kit ( Ambion ) . 2 µg of total RNA were used to perform cDNA synthesis using Superscript II Reverse Transcriptase according to the manufacturer's instructions ( Invitrogen ) . Quantitative PCR was carried out on a Mastercycler ep realplex ( Eppendorf ) with a 2X SYBR Green master mix ( SYBR Green Power , Applied Biosystems ) . The oligonucleotide primers used are listed in Table S9 in Text S1 ( oligos # 18–27 ) . The reaction mixture contained 2 . 5 µM of each primer and 5 µL of cDNA at 1∶10 , 1∶100 or 1∶1000 dilutions . Each sample was processed in triplicate . Relative expression levels were calculated using the delta-delta Ct ( ΔΔCt ) method , with C . albicans translation elongation factor CEF3 transcript as a calibrator . The relative expression was calculated as 2 ( Ct target – Ct CEF3 CEC1509 or CEC1997 ) – ( Ct target– Ct CEF3 CEC1535 or CEC2001 ) . Strains co-expressing Sfl1p-TAP and Efg1p-HA or Sfl2p-TAP and Efg1p-HA ( AVL12-SFL1-TAP or AVL12-SFL2-TAP , respectively , Table 1 ) together with the control strains SFL1-TAP , SFL2-TAP and AVL12-pHIS ( Table 1 ) were grown during 4 h in 50 ml SC medium at 30°C or Lee's medium at 37°C prior to crosslinking with formaldehyde . Cells were lysed with glass beads and total extracts were prepared in 700 µl lysis buffer ( 50 mM HEPES-KOH pH 7 . 5 , 140 mM NaCl , 1 mM EDTA , 1% Triton X-100 , 0 . 1% Na-deoxycholate ) then sonicated as described for the ChIP-Seq experiment . Immunoprecipitation was performed with 500 µl of clarified sonicated extracts and 40 µl of IgG-coated magnetic beads ( Dynabeads Pan mouse IgG , Invitrogen ) , previously prehybed overnight with PBS-0 . 1% BSA . The beads were washed once with 1 ml lysis buffer and three times with lysis buffer supplemented with 150 mM NaCl . Reverse crosslinking was achieved by incubating beads at 100°C during 25 min in reverse–crosslinking buffer ( 2% SDS , 0 . 5 M 2-mercaptoethanol , 250 mM Tris , pH 8 . 8 ) . The immunoprecipitates were resolved by electrophoresis on an 8% SDS-polyacrylamide gel . Proteins were electrophoretically transferred to nitrocellulose membranes . Blots were revealed with rat monoclonal anti-HA peroxidase conjugate - High Affinity ( clone 3F10 , Roche ) for detection of co-immunoprecipitated Efg1p-HA or with Peroxydase-Anti-Peroxydase Soluble complex ( Sigma Aldrich ) for detection of immunoprecipitated Sfl1p-TAP and Sfl2p-TAP at a 1∶2000 dilution . Gene Ontology functional enrichment analyses were conducted using the CGD Gene Ontology ( GO ) Term Finder tool ( http://www . candidagenome . org/cgi-bin/GO/goTermFinder ) . The orf19 list of the Sfl1p and Sfl2p common targets or the orf19 list of the Sfl2p-specific targets was used as input for functional grouping . To decide which of the two ORFs sharing the same bound promoter are included among the GO-term finder input list , we selected those ORFs showing differential expression in Sfl1p and Sfl2p transcriptomics data ( expression level fold-change ≥1 . 5 , P-value ≤0 . 05 ) . This led to a list of 110 ( Sfl1p and Sfl2p common targets ) and 73 ( Sfl2p specific targets ) genes for GO term enrichment analyses ( Table 2 ) . If some GO terms contained overlapping gene lists , the GO term with the largest number of genes or with the best significance score was selected . The P-value cutoff for considering a functional grouping enrichment was P≤0 . 05 . For motif discovery analyses , peak summit location files generated by the MACS algorithm [46] were imported into the Galaxy NGS analysis pipeline and DNA sequences encompassing ±250 bp around peak summits in Sfl1p or Sfl2p data sets were extracted using the Extract Genomic DNA tool version 2 . 2 . 2 . The resulting sequences were used as input for motif discovery using the SCOPE ( Suite for Computational Identification of Promoter Elements , version 2 . 1 . 0 ) program ( http://genie . dartmouth . edu/scope/ ) [56] or the Regulatory Sequence Analysis Tools ( [RSAT] http://rsat . ulb . ac . be/rsat/ ) peak-motifs algorithm [55] . The parameters used in RSAT peak-motifs algorithm were as follows: oligo-analysis and position-analysis were selected; oligo length was 6 and 7; the Markov order ( m ) of the background model for oligo-analysis was set to automatically adapt to sequence length; the number of motifs per algorithm was 10 and both strands of the DNA sequence inputs were searched for motif discovery . For building a control set of sequences ( that is sequences randomly chosen from the genome ) , we used the RSA tool “random genome fragments” . The parameters used in SCOPE were as follows: species selected was C . albicans ( genome sequence available at www . broad . mit . edu/annotation/genome/ ) ;“fixed” was selected for the upstream sequence control set and both strands of the DNA sequence inputs were searched for motif discovery . ChIP-Seq and microarray data can be found at the Gene Expression Omnibus ( http://www . ncbi . nlm . nih . gov/projects/geo/ ) or ArrayExpress ( http://www . ebi . ac . uk/arrayexpress/ ) databases under series numbers GSE42886 or E-MEXP-3779 , respectively . | Candida albicans can switch from a harmless colonizer of body organs to a life-threatening invasive pathogen . This switch is linked to the ability of C . albicans to undergo a yeast-to-filament shift induced by various cues , including temperature . Sfl1p and Sfl2p are two transcription factors required for C . albicans virulence , but antagonistically regulate morphogenesis: Sfl1p represses it , whereas Sfl2p activates it in response to temperature . We show here that Sfl1p and Sfl2p bind in vivo , via divergent motifs , to the regulatory region of a common set of targets encoding key determinants of morphogenesis and virulence and exert both activating and repressing effects on gene expression . Additionally , Sfl2p binds to specific targets , including genes essential for hyphal development . Bioinformatic analyses suggest that Sfl1p and Sfl2p control C . albicans morphogenesis by cooperating with two important regulators of filamentous growth , Efg1p and Ndt80p , a premise that was confirmed by the observation of concomitant binding of Sfl1p , Sfl2p and Efg1p to the promoter of target genes and the demonstration of direct or indirect physical association of Sfl1p and Sfl2p with Efg1p , in vivo . Our data suggest that Sfl1p and Sfl2p act as central “switch on/off” proteins to coordinate the regulation of C . albicans morphogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"expression",
"analysis",
"cellular",
"stress",
"responses",
"genetic",
"networks",
"microbiology",
"fungal",
"physiology",
"dna",
"transcription",
"genome",
"analysis",
"tools",
"model",
"organisms",
"mycology",
"gene",
"expression",
"microbial",
"pathogens",
... | 2013 | A Comprehensive Functional Portrait of Two Heat Shock Factor-Type Transcriptional Regulators Involved in Candida albicans Morphogenesis and Virulence |
Protein domain motion is often implicated in biological electron transfer , but the general significance of motion is not clear . Motion has been implicated in the transfer of electrons from human cytochrome P450 reductase ( CPR ) to all microsomal cytochrome P450s ( CYPs ) . Our hypothesis is that tight coupling of motion with enzyme chemistry can signal “ready and waiting” states for electron transfer from CPR to downstream CYPs and support vectorial electron transfer across complex redox chains . We developed a novel approach to study the time-dependence of dynamical change during catalysis that reports on the changing conformational states of CPR . FRET was linked to stopped-flow studies of electron transfer in CPR that contains donor-acceptor fluorophores on the enzyme surface . Open and closed states of CPR were correlated with key steps in the catalytic cycle which demonstrated how redox chemistry and NADPH binding drive successive opening and closing of the enzyme . Specifically , we provide evidence that reduction of the flavin moieties in CPR induces CPR opening , whereas ligand binding induces CPR closing . A dynamic reaction cycle was created in which CPR optimizes internal electron transfer between flavin cofactors by adopting closed states and signals “ready and waiting” conformations to partner CYP enzymes by adopting more open states . This complex , temporal control of enzyme motion is used to catalyze directional electron transfer from NADPH→FAD→FMN→heme , thereby facilitating all microsomal P450-catalysed reactions . Motions critical to the broader biological functions of CPR are tightly coupled to enzyme chemistry in the human NADPH-CPR-CYP redox chain . That redox chemistry alone is sufficient to drive functionally necessary , large-scale conformational change is remarkable . Rather than relying on stochastic conformational sampling , our study highlights a need for tight coupling of motion to enzyme chemistry to give vectorial electron transfer along complex redox chains .
The relationship between dynamics and the function of proteins is important . Proteins undergo a wide range of motions in terms of time ( 10−12 to >1 s ) and distance ( 10−2 to >10 Å ) scales and any of these may be significant catalytically and related directly to function [1]–[5] . Proteins exist in an equilibrium of conformational states that define a multi-dimensional free energy landscape , enabling proteins to explore high energy states [6] . Mutagenesis can induce altered landscapes leading to energy traps with consequent effects on catalytic efficiency [7] , [8] . It is in the nature of catalysis that high energy states are populated transiently during the course of an enzyme-catalyzed reaction . The ability to study these states experimentally , and to assess their impact on biological function , is a major challenge . Evidence points to a range of spatial and temporal dynamical contributions to substrate binding , product release , and chemical catalysis [9]–[11] . There is evidence supporting a role for domain motion in catalysis in the important family of diflavin oxidoreductases typified by human cytochrome P450 reductase ( CPR ) and human methionine synthase reductase ( MSR ) [12] , [13] . Pulsed Electron Electron Double Resonance ( PELDOR ) studies of both CPR and MSR indicate landscape remodeling induced by ligand binding . Domain motion in this enzyme family has also been inferred from structural studies ( crystallographic [14]–[17] and solution state [18] ) and from pressure-dependent kinetic studies of electron transfer in CPR [12] . CPR is a membrane-bound NADPH-dependent oxidoreductase that contains FAD and FMN cofactors housed in discrete redox domains separated by a flexible hinge region [15] . CPR catalyzes electron transfer from NADPH to cytochrome P450 ( CYP ) enzymes in the endoplasmic reticulum . The relative orientation of the two flavin redox domains is variable , giving rise to “open” and “closed” conformations of the enzyme as seen in crystallographic analysis of homologous wild-type and mutant forms [16] , [19] . NMR and small angle X-ray scattering studies suggest that CPR adopts a more closed conformation on coenzyme binding [18] , similar to the conformation of crystallized rat CPR in which the dimethylbenzene moieties of the FAD and FMN cofactors are juxtaposed [15] . This closed conformation is optimal for interflavin electron transfer since the short interflavin distance enhances electronic coupling . Despite this close approach , interflavin electron transfer is slow ( ∼50 s−1 ) as measured by temperature jump [20] , [21] and flash photolysis [22] time-resolved spectroscopies . These studies imply adiabatic control of electron transfer through conformational sampling [23] . This is consistent with temperature [24] , pressure [12] , and viscosity dependence [20] analysis of electron transfer kinetics , and with the multiple conformational states of human CPR seen in PELDOR studies [12] . Whilst the closed state of CPR is optimal for interflavin electron transfer , interaction with CYP enzymes requires a more open state . FMN domain residues that interact with CYP enzymes are occluded in the closed state [25] . A sequential opening and closing of CPR during the catalytic cycle is therefore proposed to facilitate internal electron transfer and subsequent transfer of electrons to CYP enzymes [16] , [26] . This proposed cycling between open and closed conformations is consistent with impaired CYP reduction by CPR containing a non-native disulphide bond that links the FAD and FMN domains and the rescue of activity following reduction of this bond [27] . Evidence for conformational cycling during CPR catalysis is largely circumstantial . A direct means of analyzing conformational variations during enzyme catalysis is required to link the kinetics ( and energy barriers ) of conformational change to the chemical ( redox ) changes that result from hydride transfer ( NADPH→FAD ) and electron transfer ( FAD→FMN ) . There are major problems to be addressed , including ( i ) identification of the “drivers” that open and close CPR; ( ii ) discrimination between electron transfer mechanisms that rely on conformational change coupled to chemical or binding events , or stochastic sampling of conformational space ( i . e . , conformational sampling mechanisms of electron transfer [28] , [29] ) ; ( iii ) whether the timescales for opening and closure support directional electron transfer from NADPH to CYP enzymes . With these key questions in mind our strategy has been to develop a direct method for analyzing the spatial and temporal properties of domain motion in human CPR using time-resolved Fluorescence Resonance Energy Transfer ( FRET ) during catalytic turnover . Our approach employs extrinsic fluorophores ( Alexa 488 and Cy 5 ) attached at different positions on the solvent exposed surface of CPR to enable spatial ( range ∼20–80 Å ) and temporal ( range ms to s ) mapping of conformational variation during stopped-flow studies of flavin reduction by NADPH . In this way , we have been able to correlate the time dependence and extent of conformational change with individual rate constants for hydride and electron transfer in CPR . Using this direct approach , we have elucidated how motions link to enzyme chemistry , identified the “drivers” of these motions , and gained important new insight into how these motions facilitate directional transfer of electrons along human microsomal P450 chains .
We generated homology models for several closed and open structures of human CPR based on X-ray crystal structures of the homologous rat CPR ( 94% sequence identity ) [15] , [16] . The fully closed structure is shown in Figure 1 . Based on these models we reasoned that CPR can bind at least two mole equivalents of an extrinsic fluorophore through a thiol linkage to cysteine residues . Figure S1 shows the absorbance spectra for the donor ( Alexa 488 ( D ) ) and acceptor ( Cy 5 ( A ) ) fluorophores attached to cysteine residues , in fluorophore-labeled CPR ( CPR-DA ) . The fluorophore and protein concentrations determined from this spectrum indicate stoichiometric attachment of the two fluorophores , giving a total fluorophore∶CPR ratio of 2∶1 . Mass spectral analysis indicates that three cysteines are labeled using our protocol , namely C228 , C472 , and C566 ( Figure S2 ) , suggesting fractional labeling of each cysteine ( see Text S1 for detailed discussion ) . C228 is located in the FMN domain and C472/C566 in the FAD domain , as shown in Figure 1 . We have not attempted to remove the multiple cysteines in the FAD domain as we wish to study the wild-type enzyme , particularly since mutagenesis may have unknown effects on the protein dynamics . We note that there is only one labeled residue in the FMN domain , C288 . From our homology models , opening of CPR results in a decreased distance between C228 in the FMN domain and C472/C566 in the FAD domain . Figure 2A shows the emission spectra of both donor labeled CPR ( CPR-D ) and CPR-DA , where the donor is excited at 495 nm . For CPR-DA , there is significant emission arising from the acceptor ( ∼670 nm ) with a corresponding decrease in the emission arising from the donor ( ∼520 nm ) compared to CPR-D . This indicates that there is efficient FRET from donor to acceptor when bound to CPR . We observed a small emission peak at ∼670 nm for the Cy 5 labeled CPR ( CPR-A ) when excited at 495 nm ( Figure S3 ) , but the relative emission is far smaller ( <∼3% ) than that attributed to FRET . If the FMN domain moves significantly relative to the rest of CPR ( as is proposed to occur following flavin reduction ) , the FRET efficiency is expected to change , manifesting as a change in the ratio of acceptor to donor emission ( A∶D ) . We were able to determine contributions to the FRET signal from inter-protein FRET as well as direct physical interaction of the extrinsic fluorophores with the flavin cofactors . A description of these control studies is given in Supporting Information ( Text S1; Figure S4 ) . We found no evidence to indicate that either of these processes contributes to the observed emission that we attribute to intraprotein FRET . We are therefore confident that the experimental setup reports on conformational change . Several studies have suggested that CPR undergoes a conformational change associated with coenzyme binding [12] , [18] , [20] , [30] . Specifically , PELDOR spectroscopy of di-semiquinoid CPR ( containing FAD semiquinone and FMN semiquinone ) has revealed that binding of NADP+ leads to formation of a more closed distribution of CPR structures compared to ligand-free di-semiquinoid enzyme [12] . It is possible to form an enzyme-coenzyme complex by incubating oxidized CPR with NADP+ . Should binding of NADP+ induce CPR closure , the distance between C228 and the cysteines in the FAD domain will increase ( as discussed above ) , resulting in poorer FRET efficiency between donor and acceptor ( i . e . , a decrease in the A∶D emission ratio ) . Figure 2B shows the resulting A∶D ratio for the emission of the CPR-DA fluorophores excited at 495 nm when titrated against NADP+ . The individual donor and acceptor emission titrations are shown in Figure S5 , normalized for the corresponding changes in fluorescence of CPR-D and CPR-A as described in Materials and Methods . This removes effects such as quenching by aromatic residues/NADP+ and FRET involving the flavin cofactors , leaving only changes attributable to FRET between the extrinsic fluorophores . From Figure 2B , the A∶D ratio decreases with increasing NADP+ concentration and saturates with a constant , KS = 1 . 6±0 . 5 mM . These data indicate that coenzyme binding induces formation of a more closed form of CPR and demonstrate that our experimental system can detect relative domain movements in CPR . By monitoring the change in fluorescence emission of the fluorophores in stopped-flow studies of flavin reduction by NADPH , we have been able to correlate the kinetics of conformational change with enzyme chemistry . We assessed the degree of photo-bleaching of the fluorophores in oxidized CPR-D ( ex 495 nm ) and CPR-A ( ex 655 nm ) . Example traces are given in Figure S6A . In each case there is a small decrease in fluorescence emission of ∼1% over 500 s . This small amount of photo-bleaching is not used to correct subsequent traces as the magnitude of the quenching is relatively small . Next , we determined if binding of NADP+ in stopped-flow studies causes a measurable change in protein conformation as demonstrated also in NADP+ titration experiments ( Figure 2B ) . The change in FRET ( CPR-DA excited at 495 nm ) was monitored on mixing oxidized and 2-electron reduced CPR-DA with a saturating concentration ( 5 mM ) of NADP+ . Example traces are given in Figure S6B–C . The observed changes in emission following mixing with NADP+ are similar to those recorded for photo-bleaching . However , we observed small shifts in the absolute magnitude of fluorescence at t = 0 for enzyme versus NADP+ mixes compared to enzyme versus buffer control mixes . This indicates a loss in the fluorescence signal of a magnitude similar to the titration study ( Figure 2B ) in the dead time of the stopped-flow instrument , consistent with fast ( <5 ms ) conformational closure of CPR . Since coenzyme-induced closure of CPR is fast , we infer that any fluorescence changes observed beyond the instrument dead time in reactions of CPR with NADPH would be related to conformational change accompanying chemical ( redox ) change in the enzyme catalytic cycle . We extracted time-resolved changes in FRET between the D and A fluorophores as flavin reduction proceeds . In this way we assessed relative conformational change associated with flavin reduction in CPR . The time-resolved FRET response is deconvoluted from other contributions to the emission response such as quenching by aromatic residues or FRET involving the flavin moieties in a similar manner as our NADP+ titration study ( Figure 2B ) . This is achieved by subtracting the traces for CPR that contained only a single fluorophore ( species CPR-D and CPR-A; fluorescence traces shown in Figure S7A and B , trace ( i ) ) from the fluorescence traces for the corresponding fluorophore in a FRET pair ( Figure S7B , trace ( ii ) ) . The resulting difference traces ( Figure 3A ) then show the fluorescence emission due to FRET between the extrinsic fluorophores alone . Opposition of the D and A traces was not observed ( as expected for FRET data ) despite deconvolution of the FRET response , suggesting we recover an approximation of the pure FRET signal . Consequently , we have not calculated detailed distance information from the FRET data , but simply used the FRET signal qualitatively to follow changes in the distribution of CPR conformations in a time-resolved manner . To extract rate constants for the observed changes in the conformational distribution , we simultaneously fit both the donor and acceptor traces in Figure 3A to a multi-exponential expression ( Text S1 ) with linked rate constants for each kinetic phase . This method is robust as the shifting sign of the amplitude for each kinetic phase facilitates good resolution of potentially similar rate constants and small amplitudes . Data fitting is described in detail in Supporting Information ( Text S1 ) . The extracted observed rate constants and amplitudes are given in Table S1 . These data can be minimally fit to a four exponential function , suggesting there are at least four conformational transitions that occur during flavin reduction by NADPH . The extracted rate constants are essentially the same for each kinetic phase for any of the traces shown in Figure S7 ( Table S1 ) . This is consistent with our assertion that the observed changes in fluorescence emission of the fluorophores are due to conformational changes in the enzyme only . That is , the traces give the same rate constants , despite different mechanisms ( quenching , FRET , etc . ) , since the changes in fluorescence emission are caused by the same conformational change as flavin reduction proceeds . Moreover , changes in the FRET signal report specifically on distance changes between C228 and C472/C556 as shown by control experiments in which negligible changes in A∶D ratio were seen with a variant form of CPR containing the C228S mutation ( see Text S1 and Figures S8 and S9 ) . Absorption studies of flavin reduction by NADPH in CPR using stopped-flow methods have previously been reported [12] , [24] , [31] and can be used to dissect flavin reduction in CPR in detail . These studies indicate that NADPH binds to the FAD domain where it transfers a hydride to the N5 of FAD followed by electron transfer from FAD to FMN to yield a distribution of 2-electron reduced species ( FADH• FMNH• , FADH2 FMN and FAD FMNH2 ) . In the absence of an electron acceptor ( such as CYP ) a second equivalent of NADPH binds to the FAD domain and transfers a hydride to FAD , driving the equilibrium distribution of enzyme states towards the fully ( 4-electron ) reduced species ( FADH2 FMNH2 ) . The observed rate constants for formation of 2-electron ( FMNH• FADH• ) and 4-electron ( FMNH2 FADH2 ) reduced CPR can be monitored by following the formation and decay of the di-semiquinoid ( FMNH• FADH• ) 2-electron reduced species at 600 nm on mixing with a saturating concentration of NADPH in a stopped-flow instrument [31] . The two exponential phases extracted from these reaction traces correspond broadly to the observed rate constants for 2-electron and 4-electron reduction ( termed k1 and k2 , respectively ) . CPR-DA reacts with NADPH in a similar way , and the kinetics of absorption change at 600 nm for the two exponential phases are identified as “flavin reduction” ( black trace ) in Figure 3B . Further , there is a very slow phase ( increase in 600 nm absorbance; represented as “EQ” in Figure 3B ) observed after 4-electron reduction . In wild-type CPR , this phase has been attributed previously to the formation of an internal equilibrium between redox states in the absence of an electron acceptor [31] . This slow adjustment to the final equilibrium of redox states is retained in CPR-DA . In the present study , we focus on the chemical steps , k1 and k2 . Figure 3B shows a typical reaction trace for flavin reduction by NADPH in CPR monitored at 600 nm . The data are fit to a four exponential function ( Text S1 , Equation S1 ) accounting for all the observed absorption changes discussed above . The rate constants for k1 and k2 at 25°C are given in Table 1 , extracted as kobs1 = 16 . 2±0 . 2 s−1 and kobs2 = 4 . 0±0 . 1 s−1 . Table 1 also shows the observed rate constants for the first two kinetic phases extracted from the fluorescence data that represent conformational change ( Figure 3A ) . The rate constants extracted from the fluorescence data ( kobs1 = 24 . 5±1 . 0 s−1 and kobs2 = 4 . 2±0 . 1 s−1 ) are similar to the observed rate constants for flavin reduction , suggesting that flavin reduction and conformational change are linked in CPR . We now correlate the observed changes in flavin redox state with the conformational changes extracted from our fluorescence data . Figure 3B shows the ratio of acceptor to donor ( A∶D ) emission extracted from the stopped-flow traces shown in Figure 3A , effectively describing the trend in FRET efficiency during flavin reduction by NADPH . These data clearly show that the FRET signal is increased ( i . e . , more “open” conformations are populated ) as CPR is sequentially reduced to the 2-electron and then 4-electron levels ( indicated by the time-resolved absorption measurements at 600 nm; Figure 3B ) . Further , following flavin reduction we observed a gradual closing of CPR ( reduced A∶D emission ratio ) over prolonged time periods ( 10 to 1 , 000 s ) as CPR relaxes to the final equilibrium position . These FRET data indicate , therefore , that conformational closure is a key part of this long-time base equilibration of the reduced enzyme species and that the more open state is a metastable form of reduced CPR . The fluorescence and absorption changes shown in Figure 3B occur over very similar timescales ( 0–10 s ) , suggesting that domain motion is linked to redox chemistry . The potential for direct coupling of redox change with conformational opening of CPR is addressed below . Strong evidence for the coupling of conformational change with enzyme chemistry would arise if the energetic barriers for electron transfer and motion are shown to be equivalent . This was addressed experimentally by monitoring the temperature dependence of the rate of flavin reduction ( absorption change at 600 nm , reporting on k1 and k2 ) and associated conformational change ( from time-resolved fluorescence data ) . The temperature dependence of the rate constants k1 and k2 for structural change and flavin reduction is shown in Figure 4 . The Eyring plots for the fluorescence data are linear ( Figure 4 ) , suggesting that the first two exponential phases each report on rate constants for a single process ( i . e . , structural change ) . The values of ΔH‡ for flavin reduction and conformational change are given in Table 1 and are similar for both kinetic phases ( k1 and k2 ) . That ΔH‡ is very similar for both flavin reduction and structural change for both exponential phases ( 2-electron and 4-electron reduction ) suggests a tight coupling of the reaction chemistry with the observed structural transitions . The kinetics and energetics of flavin reduction and conformational change are consistent with the two processes being tightly coupled . These data might suggest that flavin reduction is responsible for conformational opening of CPR or that conformational change induces electron transfer associated with flavin reduction . The analysis , however , is complicated by the opposing effects of coenzyme binding , which in the absence of redox chemistry is known to effect closure of CPR ( Figure 2B ) . We therefore conducted stopped-flow measurements in which CPR-DA was mixed with the chemical reductant , sodium dithionite , to investigate the effects of redox change in the absence of coenzyme binding . A typical trace from these experiments is shown in Figure 5A . As with NADPH reduction , when dithionite is used to reduce the flavin centers the appearance and subsequent disappearance of the flavin semiquinones is observed at 600 nm corresponding to formation of the 2-electron and 4-electron reduced species of CPR . The individual exponential components corresponding to 2-electron and 4-electron reduction are less well defined , which complicates data analysis , but approximate rate constants can be obtained . The observed rate constant for flavin reduction is far slower , with dithionite being ∼0 . 05 s−1 and ∼0 . 04 s−1 , compared to NADPH , ∼19 s−1 and ∼2 . 5 s−1 , at 20°C for 2- and 4-electron reduction , respectively . The change in FRET efficiency for CPR-DA was also monitored as flavin reduction proceeds ( Figure 5A ) . Individual fluorescence traces used to calculate the FRET response are shown in Figure S11A , B . The change in FRET efficiency can be adequately fit to a two-exponential function ( Text S1 , Equation S1 ) with observed rate constants of ∼0 . 05 s−1 and ∼0 . 03 s−1 for the first and second kinetic phases , respectively . In general , there is a large increase in A∶D ( CPR opening ) as flavin reduction proceeds . We note that the first exponential phase shows a slight decrease in A∶D , though this is on a faster timescale than either 2- or 4-electron reduction ( Figure 5A ) . Therefore , as seen with NADPH , the rate of flavin reduction by dithionite correlates with the observed rate of conformational change ( Figure 5 ) and reduction to the 2-electron and 4-electron levels is accompanied by an opening of CPR . These data indicate that reduction of the flavin cofactors alone is sufficient to induce conformational opening of CPR . Further , these data are consistent with our temperature dependence data from which we inferred a tight coupling of the conformational transition with the flavin redox state . We also examined the effects of coenzyme binding on the rate of flavin reduction and conformational opening during enzyme reduction with dithionite . This is achieved by mixing CPR-DA that had been pre-incubated with a saturating concentration of NADP+ ( 5 mM ) with dithionite . There was no evidence for reduction of NADP+ to NADPH ( monitored by absorption changes at 340 nm ) by dithionite over the timescale of the study . In the presence of NADP+ , the rate constant for flavin reduction by dithionite ( 2-electron reduction ) increases approximately 2-fold ( ∼0 . 01 s−1; Figure 5B ) compared with reactions performed in the absence of NADP+ . The conversion of 2-electron reduced CPR to the 4-electron reduced species is not well-resolved , due likely to overlap with the slow kinetic phase ( s ) involved in establishing the final equilibrium of redox states ( EQ ) ( Figure 5B ) . The corresponding change in FRET efficiency is shown in Figure 5B with the individual fluorescence trace shown in Figure S11C and S11D . As with dithionite alone , there is a significant increase in the A∶D ratio corresponding to CPR opening and this occurs on a similar timescale to flavin reduction ( approximate rate constant ∼0 . 08 s−1 ) . The effect of NADP+ is therefore to accelerate ( approximately 2-fold ) flavin reduction and the associated conformational opening of CPR with dithionite as reductant . Further , after the opening of CPR with NADP+ bound there is a subsequent decrease in A∶D reflecting CPR closure as the reduced enzyme relaxes to the final equilibrium ( EQ ) state ( Figure 5B ) . On the timescale of our measurements we do not observe the establishment of this equilibrium in dithionite studies performed in the absence of NADP+ ( Figure 5A ) . We find therefore that NADP+ not only increases the observed rate of flavin reduction , but also increases the observed rate of EQ formation . This is consistent with previous t-jump studies of interflavin electron transfer in di-semiquinoid human CPR , where the observed rate of electron transfer ( 55 s−1 ) is increased 5-fold on adding NADP+ compared to reactions performed in the absence of nicotinamide coenzyme [20] . The precise reasons for accelerated flavin reduction in the presence of NADP+ are unclear . However , cofactor binding likely induces a shift in the equilibrium distribution of enzyme forms towards a more closed conformation ( Figure 2B ) . We suggest that internal electron exchange between the flavin cofactors is enhanced due to a minimized cofactor separation induced by NADP+ binding . Clearly , once CPR is reduced by dithionite the distribution then adjusts first to the metastable open conformations ( 0–50 s ) and then relaxes to the more closed EQ conformations ( >50 s ) ( Figure 5B ) . We have monitored two separate conformational transitions in CPR , namely opening and closing , which correspond to increased and decreased separation of the FMN and FAD cofactors , respectively . Opening is driven by reduction of the flavin cofactors and closure by coenzyme binding ( Figure 6A ) . We now develop an integrated model for CPR action that incorporates these conformational transitions ( Figure 6 ) . In this model , domain motions driven by flavin reduction are crucial in mediating electron transfer to CYPs . The open conformations expose residues required for CYP interaction that are occluded in the closed conformation [25] . The open state signals that CPR is “ready and waiting” to transfer electrons to CYP partners in the microsomal membranes . Should productive interaction with CYP partners not occur , subsequent closure of CPR ( formation of EQ ) may offer some protection of the reducing equivalents in the flavin cofactors by suppressing their adventitious transfer ( e . g . , to molecular oxygen ) . The open conformation of CPR is not appropriate for “electron loading” from the reducing coenzyme NADPH . Rapid equilibration of electrons across the flavin centers is required to generate 2- and 4-electron reduced CPR from the oxidized form . Therefore , “closure” of the oxidized form of CPR is induced on nicotinamide coenzyme binding to facilitate efficient loading with reducing equivalents prior to redox-driven opening of the structure . We note that there is evidence for both 2 , 4 and 1 , 3 electron cycling in CPR [32] , [33] , and we propose a similar mechanism of opening/closing driven by the flavin redox state can occur in either case . However , in vivo the need for a second hydride transfer from NADPH may be less important ( Figure 6B ) . The redox and ligand-bound forms of CPR therefore drive the re-distribution of CPR conformations across the associated energy landscape to more open or closed forms of CPR . These different conformations direct downstream interaction of CPR with CYP partners and facilitate directional transfer of reducing equivalents for CYP-mediated catalysis . Such motion therefore drives the vectorial transfer of electrons from NADPH to CYP to catalyze the wide range of mono-oxygenation reactions in the endoplasmic reticulum . It is important to distinguish between the relatively large-scale redox-coupled and ligand-coupled motions discussed above and other stochastic motions that can limit the rate of electron transfer . Our model therefore also recognizes that smaller-scale motions can also limit electron transfer , either between flavin cofactors ( in the closed state ) or to CYP enzymes ( in the open state ) . Localized searches for productive reaction geometries are common in biological electron transfer reactions and these are often responsible for the slower observed rates of electron transfer compared to those predicted for “pure” ( nonadiabatic ) electron transfers on the basis of distance criteria alone [23] , [34] . Indeed , our temperature-dependence studies indicate that the reaction cannot be modeled using the Marcus nonadiabatic formalism for electron transfer ( see Text S1 for details; Figure S10 and Table S2 ) . As such , Figure 6 should not be taken to imply that single discrete open and closed conformations of CPR exist under a defined ligand-bound or redox form of the enzyme . Rather , an ensemble of conformations exist ( as indicated by PELDOR studies of different liganded forms of di-semiquinoid CPR [12] ) and that redox change and ligand binding drive the equilibrium distribution towards more open or closed states , respectively . We suggest that localized searches for reactive electron transfer geometries could be rate limiting for interflavin electron transfer in CPR , consistent with the slow observed rates of electron transfer in t-jump [20] and laser flash photolysis experiments [20] , [35] . The use of time-resolved FRET in conjunction with stopped-flow absorption analysis of the CPR catalytic cycle has enabled us to present a dynamic model of catalysis in which redox change and ligand binding drive large-scale redox domain motion . By coupling conformational change to redox change and ligand binding , CPR optimizes internal electron movements between the flavin cofactors and signals “open and ready” conformations to partner CYP P450 enzymes . This linking of motion with enzyme chemistry enables fine control of vectorial electron transfer along the NADPH→FAD→FMN→heme ( CYP ) chain that supports all P450-mediated catalysis in the microsome . Given the structural similarity of CPR with other major mammalian diflavin oxidoreductases , including the isoforms of nitric oxide synthase , we anticipate CPR will be a prototype for similar coupling of reaction chemistry , ligand binding , and motions in biology .
The human CPR structures were modeled using SwissModel from the rat CPR crystal structures . For the closed and intermediate forms , the sequence was simply fit to the crystal structure ( 1AMO_A and 3ES9_C , respectively , with 94% and 93% sequence identity ) . Loops not present in the crystal structures were modeled using the built-in loop database ( closed form: residue 235–241 and 499–505; intermediate form: 235–238 and 499–504 ) . For the most open form , a significant portion of the FMN domain is missing in the crystal structure ( 3ES9_B ) . This portion was modeled by aligning the FMN domain from the intermediate form onto the existing coordinates of the crystal structure with the loop connecting the two domains modeled using the built-in loop database . Human CPR was expressed and purified essentially as described previously [31] . Labeling of CPR with extrinsic fluorophores was achieved by incubating CPR in 50 mM potassium phosphate , pH 7 at <20°C in an anaerobic glove box ( Belle Technology ) with either Alexa 488 C5-maleimide ( Molecular Probes ) or Cy 5 mono-maleimide ( GE Healthcare ) . To achieve a 1∶1 ratio of Alexa 488 and Cy 5 bound to CPR , incubation was with 1 mM and 5 mM of the fluorophores , respectively . Non-reacted fluorophore was separated from the sample by running through a desalting column equilibrated with 50 mM potassium phosphate , pH 7 . Details of mass spectral analysis are given in Supporting Information ( Text S1 ) . Unless otherwise stated , CPR was fully oxidized prior to all experiments by adding a few grains of potassium ferricyanide and elution through a desalting column as above . 2-electron reduced CPR was formed by reaction with an equimolar concentration of NADPH ( Melford ) , and elution through a desalting column under anaerobic conditions . Fluorescence emission spectra were monitored on a Varian Cary Eclipse fluorescence spectrophotometer ( Varian Inc . , Palo Alto , CA , USA ) . Multiple wavelength absorbance spectra were monitored on a Varian Cary 50 Bio UV/Vis spectrophotometer . Specific experimental conditions are given in the main text . All experiments were performed in 50 mM potassium phosphate , pH 7 . The saturation constant , KS , was extracted by fitting concentration dependence data to a weak binding function ( Equation 1 ) : ( 1 ) To prevent oxidase activity of CPR , all kinetic studies were performed under strict anaerobic conditions within a glove box ( Belle Technology; <5 ppm O2 ) using a Hi-Tech Scientific ( TgK Scientific , Bradford on Avon , UK ) stopped-flow spectrophotometer housed inside the glove box . Spectral changes accompanying flavin reduction and flavin di-semiquinone formation/decay were monitored at 456 nm and 600 nm , respectively . For reduction with sodium dithionite , the same solution of dithionite was used for all experiments within 6 h . Fluorescence emission changes associated with donor emission were monitored using a 550 nm short wave pass optical filter . Fluorescence emission changes associated with acceptor emission were monitored using a 650 nm long wave pass optical filter . For the given reaction conditions , no bleed through of fluorescence was observed . Data were collected over a log timebase ( 15 decades , 3 , 000 data points total ) . Typically 3–5 measurements were taken for each reaction condition . Fitting of reaction traces is described in detail in Supporting Information ( Text S1 ) . | Enzymes are proteins that catalyze a large array of chemical reactions , often in partnership with other enzymes . We understand in detail the chemical mechanisms of many of these reactions; however , the importance of the physical movements of enzymes during catalysis ( or protein dynamics ) is , increasingly , becoming apparent . In this study , we have placed fluorescent markers on an enzyme called cytochrome P450 reductase ( CPR ) to probe the dynamic changes in the physical conformation of the protein as the reaction chemistry proceeds . CPR catalyses the transfer of electrons from a small molecule donor ( called NADPH ) , ultimately passing them to their partner enzymes called CYPs . We were able to correlate specific conformational changes with distinct chemical steps in CPR . We found that the chemical transformation itself induces the enzyme to adopt conformations that are required for its efficient interaction with CYPs . These findings have allowed us to develop a model of CPR activity in which electron transfer along the pathway from NADPH through CPR to CYP is tightly integrated with physical conformational control of the enzyme . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"biology",
"biophysics"
] | 2011 | Coupled Motions Direct Electrons along Human Microsomal P450 Chains |
A key challenge in movement ecology is to understand how animals move in nature . Previous studies have predicted that animals should perform a special class of random walks , called Lévy walk , to obtain more targets . However , some empirical studies did not support this hypothesis , and the relationship between search strategy and ecological factors is still unclear . We focused on ecological factors , such as predation risk , and analyzed whether Lévy walk may not be favored . It was remarkable that the ecological factors often altered an optimal search strategy from Lévy walk to Brownian walk , depending on the speed of the predator’s movement , density of predators , etc . This occurred because higher target encounter rates simultaneously led searchers to higher predation risks . Our findings indicate that animals may not perform Lévy walks often , and we suggest that it is crucial to consider the ecological context for evaluating the search strategy performed by animals in the field .
How should we move to search for targets when we have no information about their location ? This is called the random search problem , which has attracted the attention of researchers in various fields [1] . The problem can be applied to various phenomena , including molecular-level movements within an organism , cell movements , movements of an individual animal , and the movement of robots [2–4] . For example , animals search their environment for food , prey , mates , and nesting locations , and DNA-binding proteins move around to find a specific DNA sequence to initiate gene expression . The search strategy is considered to evolve to be more efficient through the process of natural selection because successful searches increase fitness , especially at the individual level in animals . The Lévy walk search ( or foraging ) hypothesis was proposed to solve the random search problem [5] . A Lévy walk is a special class of random walk models in which the probability function of step length l has a power-law tail: P ( l ) ∼l−μ ( 1<μ≤3 ) , where μ is a power-law exponent , such that rare ballistic movements occur among a number of relatively short steps . Comparisons of the efficiency of random searches showed that a Lévy walk with μ ≈ 2 was a highly efficient search strategy in environments where patchy prey were sparsely distributed [1 , 5–7] . In dense environments , on the other hand , Lévy walks had almost the same efficiency as Brownian walks [6] . Therefore , the Lévy walk foraging hypothesis predicts that most animals should perform Lévy walks while searching unless there are abundant targets . Although many empirical studies have reported that diverse organismal components and taxa ( e . g . , T cells , insects , and human beings ) perform Lévy walks with μ ≈ 2 [1 , 3 , 8–15] , several recent analyses demonstrated that some animals had various Lévy exponents μ , or they exhibited Brownian walks [13 , 14 , 16–18] . For example , rigorous statistical analyses of deer and bumblebees failed to provide strong evidence for Lévy walks [16] . Thus , the question changed from whether animals have Lévy walk movement patterns to when or why animals perform Lévy walks . In general , the diversity of organisms is the result of varying ecological and environmental factors as well as complex biotic interactions with conspecific and heterospecific individuals [19] . Theoretical reports of random searches have generally focused only on search efficiency to evaluate the fitness of the searcher [1 , 5–7 , 20–23] . Moreover , most of these studies paid little attention to other relevant ecological factors such as death rate with predation risk , interactions with other individuals , and the metabolic costs of foraging . A few studies have considered such factors [24–28] , including predation risk [29–32] . In physics , Yuste et al . analyzed the survival rate of mortal random ( Brownian ) walkers surrounded by diffusing traps and revealed that the death in the course of motion dramatically affected the search efficiency [33] . Such a situation would be relevant to biological encounters . However , the relationships between search efficiency and predation risks in the context of Lévy walks are still poorly understood . Here , we focus on the fact that search efficiency represents the probability of an encounter with anything existing in the environment . Highly efficient search strategies may correspond to more frequent encounters with predators , and thus higher death rates . Therefore , search efficiency , as defined in previous studies , may not reflect the actual fitness , because fitness is not only determined by the efficiency of searching for targets ( i . e . , benefits ) , but also by the death rate caused by predation ( i . e . , cost ) [34] . Reynolds [30 , 31] reported that predation risk altered the optimal strategy , but did not consider predators [30] or the fitness of the searcher [31] . In this paper , we explicitly introduce predation risk and life-cycle types to the previous simulations , and extend the random search scenario to correctly estimate the fitness of a searcher to determine an animal’s optimal search strategy .
First , we considered a searcher performing either the Lévy walk ( hereafter , LW ) or the Brownian walk ( hereafter , BW ) at movement velocity vs ( = 1 ) in an environment in which patchy targets were sparsely distributed . Then , Np predators were randomly placed in the environment as the initial condition . To explore the effect of the predators’ movements , we considered four cases with respect to the predators’ movement velocity vp/vs = 0 ( sit-and-wait ) ; vp/vs = 0 . 2 ( slow ) ; vp / vs = 1 ( middle ) ; and vp / vs = 5 ( fast ) . If vp > 0 , we assumed that a predator performed LW with μ = 2 ( or BW in S1 Text ) . For simplicity , we assumed that if the searcher encountered a predator , the searcher died from predation . Second , if the death effect arising from encounters with predators was considered , search time became an important factor because the length of rest during searches was associated with fitness . Thus , each searcher had a maximum searching time , Tmax , that could be cut off by an encounter with a predator . Finally , when considering the searcher’s fitness , we assume: ( 1 ) the fitness is the lifetime reproductive success ( i . e . , we analyze the number of offspring reproduced within the lifetime ) , ( 2 ) without alternation of generations ( i . e . , we do not take population dynamics into account ) , ( 3 ) a searcher has either one of two life-cycle types as described next . In the simplest case , finding a target directly led to increased fitness in a linear fashion ( life-cycle type I ) . For example , when a female parasitoid wasp finds and attacks a host , and then searches for another target , we presume that its fitness increases linearly . Furthermore , when a male finds a female and mates , its fitness as a searcher also increases linearly . In contrast , animals characterized by life-cycle type II would need to survive until their reproductive stage Tmax to obtain higher fitness . In life-cycle type II , individuals that die from predation prior to maturity have no offspring and have a fitness of zero . Here , we show the general relationship between fitness and the rate of encounter with targets and predators as well as the robustness of Tmax to our results . We denote the encounter rate with a predator per unit time ΔT as γ . The probability of an encounter with the predator at the m-th time unit is expressed as ( 1−γ ) m−1γ . ( 1 ) Therefore , when Tmax is divided into n pieces by unit time ΔT ( i . e . , Tmax = ΔTn ) , the mean search time T¯ is T¯=∑m=1n{m ( 1−γ ) m−1γ}+n ( 1−γ ) n=1− ( 1−γ ) nγ , ( 2 ) where n ( 1−γ ) n indicates the case in which the searcher never encounters predators . When the mean number of encounters with predators for n is k , nγ = k and ( 1−γ ) n ≈ e−k for γ << 1 and a large n , thus T¯≈n ( 1−e−k ) k . ( 3 ) We calculated the fitness of LW and BW strategies in the ecological context using computer simulations because it is difficult to analytically derive the encounter rate in our relatively complicated setting , even though the analytical solutions were obtained in different scenarios under much simpler assumptions ( e . g . , Brownian walks , 1-D field , ideal gas model ) [5 , 33 , 35 , 36 , 37] . Using the methods described previously [5 , 7 , 17] , we simulated one searcher roaming in a 2-D environment in which some targets ( e . g . , food , hosts , mates ) and predators were distributed . Although the species at higher trophic levels are lower in number in real ecosystems and the population we simulated seems unsustainable , we introduced only one searcher . This is because we focused on the fitness of a single searcher by picking it up from searcher’s population , and our main results must be robust if we introduce a number of searchers . The searcher had no prior information about the locations of both targets and predators , and wandered at a constant velocity vs = 1 ( per unit time ) in a 2-D continuous field with length L2 = 500×500 in which the boundary condition is periodic [7] . The LW was characterized by a distribution function P ( l ) ∼ l−μ ( 1 < μ ≤ 3 ) . In our simulations , we derived step lengths from the following equation to obtain LW , generating a uniform random number u ( 0 < u ≤ 1; except for u = 0 ) : l=l0u ( 1−μ ) −1 , ( 7 ) where the minimum step length l0 is 1 [7] . For the BW simulation , to obtain an exponentially decaying distribution of the move length , each successive step length was drawn from a Gaussian distribution , where the mean was the minimum step length l0 = 1 and the variance was equal to 1 [7] . In LW or BW , after walking in a straight-line motion until reaching a step length l , the searcher turns in the angle drawn from a uniform distribution [−π , π] . The center of each patch was randomly scattered , and the radius of each patch was equal to 10 . The number of targets and patches in the whole field was 1000 and 50 , respectively . Each target was randomly assigned to a patch so that each patch had 20 targets on average . The targets were randomly distributed within the patch . In the initial state , Np predators were randomly distributed in the whole area , i . e . , the x and y position of each predator was independently drawn from a uniform distribution [0 , L] ( See S1 Text for the effects of initial conditions ) . Rt and Rs represented the radius of the targets and searcher , respectively , and Rs′ and Rp′ represented the radius of perception of the searcher and predators . If the distance between the searcher and a target was less than Rt+Rs′=1 , the searcher obtained the target , and the target disappeared . Then the step length of the searcher is truncated and recalculated , and the direction is drawn from a uniform distribution . After the searcher migrated a 500 path length , the depleted target regenerated to maintain the specified target density [17] . Similarly , the searcher died if the distance between the searcher and a predator was less than Rs+Rp′=1 . The mean free path λ , which represents the mean distance or travel time between patches or targets , is L22RN for the 2-D environment [7] . Hence , in our simulation , λpatch = 2500 and λtarget = 125 for encounter distance R = 1 . This is equivalent to the low-resource scenario of previous studies ( e . g . , [22] ) . The maximum search time Tmax was 104 . To converge the results , the total time for a single parameter set ( i . e . , searcher’s movement pattern and density of predators ) was 107 for sit-and-wait , slow , or middle predator conditions , and 5×107 for fast conditions . Then , k , η , γ were calculated , and the relative fitness was obtained using Eqs ( 5 ) and ( 6 ) .
The results of the relative fitness ( ϕLW / ϕBW ) calculation for life-cycle type I are presented in Fig 1 . When predators were absent ( Np = 0 ) , the relative fitness ϕLW / ϕBW was >2 . Thus , LW with intermediate-level μ had the highest fitness , which was consistent with the findings of previous studies [5 , 7] . However , as the number of predators increased , the ϕLW / ϕBW ratio gradually declined to ~1 or slightly less than 1 when the predator strategy was sit-and-wait or slow LW ( Fig 1A and 1B ) . In this case , a LW ( μ ≈ 2 ) often performs best out of all LW’s . When the predator strategy was middle or fast LW , ϕLW / ϕBW was maintained at a high value , and LW could be an efficient strategy ( Fig 1C and 1D ) . Likewise , in the case of a searcher with life-cycle type II , the relative fitness ϕLW/ϕBW decreased substantially as the number of predators increased when the predator strategy was sit-and-wait or slow LW ( Fig 2A and 2B ) . Even when the strategy of predators was middle LW , ϕLW/ϕBW decreased as the number of predators increased . These results were robust to other search strategies ( i . e . , correlated random walk or composite Brownian walk ) ( S1–S4 Figs ) and to Brownian walk predators ( S5 Fig ) . The relative fitness decreased because the searcher was likely to encounter a predator . The search time was shortened by death in a manner dependent on the search efficiency , and the relative mean searching time T¯LW/T¯BW depended on the search strategy ( Fig 3 ) . These results indicated that the LW strategy could lead to a high predator-encounter rate; therefore , BW could potentially be a risk-averting strategy . To investigate these results , the relationship between the relative fitness and the encounter rate with targets and predators was examined ( Fig 4 ) . This result is not limited to our simulation results or to the relative fitness of LW or BW , but it describes a general trend . The relative encounter rates with targets and predators and the expected encounter number of BW for our simulation are presented in Fig 5 . When the encounter rate with predators was low ( i . e . , low kBW ) , the fitness of random search strategies clearly depended on the encounter rate with targets ( Fig 4A and 4D ) . Hence , LW had higher fitness in our simulation ( Fig 5A ) . On the other hand , for intermediate or high kBW , fitness also changed depending on the encounter rate with predators ( Fig 4B , 4C , 4E and 4F ) . Furthermore , fast predators displayed the same high predator encounter rates of high kBW and γLWγBW≈1 ( Fig 5B and 5C ) . Thus , LW had higher fitness under the fast-predator conditions for life-cycle type I and almost equal fitness for life-cycle type II . Similarly , the fitness of other random search strategies was determined by the encounter rate with targets and predators . The degree of encounter rate improvement not only depends on the search strategy , but also on the distribution or density of the targets [5 , 7] , suggesting that the conditions for the optimal search strategy are complex .
Our results revealed that the random search strategy affected the death rate arising from predation , and that trade-offs could occur between foraging efficiency and predation risk . In nature , animal species have different ecological traits or interactions associated with their foraging behavior [34 , 38 , 39] . Considering such ecological factors , optimal foraging theory , as it currently exists , successfully predicts various types of animal behaviors from the viewpoint of maximizing fitness through natural selection [38 , 39] . However , previous studies of random search movements have only focused on foraging ( i . e . , search efficiency for targets ) , which may be unrealistic when considering the diversity of ecological characteristics and biotic interactions in nature . Lima et al . [34] reported that animals performed more efficient strategies in response to ecological factors , including risks , with such trade-offs . Our simulations predicted that where predators were abundant , a searcher performing a LW might have lower fitness depending on its ecological characteristics and those of the predators . This suggests that the optimal search strategy may change . Therefore , the parameter range in which the LW is advantageous may be narrower than previously estimated ( Fig 4 ) . The mechanism explaining these dynamics was that LWs not only increased the encounter rate with targets , but also with predators , which shortened the lifespan in exchange for the capture of more targets . The rare ballistic movements of LWs led to the high encounter rate with predators ( Fig 5C ) , and this effect has been reported as a high encounter rate of a straight line motion with randomly distributed destructive targets [7] or new targets [35] . In the presence of predators , a searcher was confronted with conditions similar to the destructive search problem , because encounters with predators resulted in the death of the searcher . Although we assume the ecological context in this paper , such searching-avoiding trade-offs in the random search problem that we revealed here may occur in other contexts such as protein-DNA interactions [2 , 40] . Previous studies analyzing the predation effect on search strategies focused on the predation risk within a patch [30 , 31] , and reported that the predation risk could alter the optimal time spent for intensive searches if the predation risk increased as the time spent within a patch increased . In contrast , we concentrated on the predation risk in a whole area and predicted the fitness ratio between LW and BW by calculating the encounter rate with targets and predators . Also , Reynolds simulated the moving preys searched by one predator [31] , and the study discussed that the prey movement patterns were determined by their foraging and not by cost of predation when predators are fast . This idea is consistent with our results for life-cycle type I ( especially in Fig 1D ) , but we defined the fitness based on life-cycle and simulated the tri-trophic system consisting of targets , searchers , and predators . Consequently , we revealed the general effect of predation risk on search strategy ( Fig 4 ) . To disentangle the effects of density , radius , and velocity of a searcher or predators on the relative fitness , we refer to analytical results of simple situations . Hutchinson et al . [35] and Dusenbery [37] reviewed the analytical results for the encounter rate of two kinds of straight motion agents ( e . g . , target and searcher , or searcher and predator ) in 2-D and 3-D . In this case , the encounter rate is proportional to both the density of agents and encounter distance ( i . e . , Rs+Rp′ in our model ) . In our results , the density of predators is an important factor that can determine the relative fitness . In Fig 5C , the left ( low density ) and right ( high density ) figures are almost identical because the effect of predator’s density in the ratio of encounter rate γLWγBW is cancelled out . Although the movements in our simulation are not straight motions but LW or BW , the proportionality of density effects on encounter rate could be common . Therefore , the density of predators can affect only the number of encounters to predators k . The ratio of encounter rate depends on the characteristic of movements ( i . e . , LW or BW , and velocity ) rather than the density of predators . Additionally , the radius of searcher and predators , that is , encounter distance can be also the same effect as the density of predators in our results because the encounter rate can be proportional to the encounter distance . The encounter rate in 2-D of a stationary searcher and straight motion predators with constant speed vp is 2ρRsvp where ρ is the density of predators [37] , and that of a straight motion searcher and predators with the speed vs = vp is 8ρRsvp / π [35] . Hence , the ratio of the encounter rates is 4 / π . This is consistent to our result for the ratio of encounter rate of BW ( i . e . , like a stationary searcher ) and LW with small μ ( i . e . , like a straight motion searcher ) under the presence of LW predators ( i . e . , like straight motion predators ) in the case of vp/vs = 1 ( green line in Fig 5C ) . In the case of vp/vs = 0 , 0 . 2 , 5 , the movement of the faster individuals has a large effect on the encounter rate [37] . Therefore , compared with BW , LW in vp/vs = 0 , 0 . 2 has the high encounter rate with predators ( Fig 5C ) . Additionally , the analytical result for 3-D conditions is similar to that for 2-D [37] . Thus , our conclusion could be applied to 3-D such as prey-predator interactions of planktons in lakes or ocean . Moreover , a recent study proposed a framework for encounter rates that are derived from an arbitrary trajectory of a searcher and immobile targets using an encounter kernel [41] . The combination of this technique and our results for general relationship between encounter rate and fitness ( Fig 4 ) may provide the general framework integrating movements and fitness . This could give us the information about fitness directly from the trajectory and distribution of targets . Many empirical studies have reported that the movement patterns of animals , from insects to human beings , are expressed as LWs with μ ≈ 2 [1] . However , the power-law exponents fitted to movement patterns sometimes ranged from 2 to 3 [1] , suggesting that movement patterns may be diverse . Additionally , the data best fitted to the exponential decay distribution ( i . e . , BWs ) has also been reported [13 , 14 , 16 , 18 , 42 , 43] . In theoretical studies , the first attempt reported that LWs with μ ≈ 2 were optimal for targets that can be revisited ( i . e . , non-destructive ) or those that are extremely patchy [5] . Moreover , LWs with μ→1 ( i . e . , straight movement ) were the optimum for randomly distributed destructive targets . After the study , the results of several versions of simulations suggested that LWs with 1 < μ ≤ 2 are more efficient depending on the prey distribution and other factors [20 , 21 , 26] . For the power-law exponent μ > 3 ( i . e . , BW ) , it has been theoretically reported that the foraging efficiency is similar to LW under high-resource conditions [6] . Our results suggests that under high predation risk , animals with power-law exponents close to three have higher fitness than μ ≈ 2 or μ < 2 ( Figs 1 and 2 ) , and those under intermediate predation risk , LW with 2 < μ < 3 also benefit . Therefore , it can be an alternative explanation for the diversity of power-law exponents . There is a question of whether movements in animals are spontaneous patterns for adaptation or a reflection of interactions with targets or complex environments [42] . de Jager et al . experimentally explained Brownian movement patterns of mussels by truncations resulted from encounters with conspecific individuals , which is the original mechanism of Einstein’s collision-induced BWs [42 , 44] . In contrast , our findings suggested that spontaneous BWs were beneficial , and this conclusion is supported by the fact that the pattern can spontaneously change depending on internal physiological states [45 , 46] . Of course , our hypothesis does not contradict the claim of de Jager et al . , because the spontaneous LW pattern has higher efficiency in the absence of risk . Furthermore , our results suggest that animals can change their search strategy according to their developmental stage or in response to predator cues . For example , a juvenile individual under high predation pressure might adopt the BW strategy to avoid predator encounters , but an adult might adopt the LW strategy to obtain more targets in the absence of predators or under low predation pressure . In smaller scale responses , when an individual receives a chemical cue ( kairomone ) that indicates the presence of a predator , switching the internal pattern from LW to BW may represent an adaptive searching strategy , because the stochastic or random pattern can arise from internal processes [32 , 46–49] . Although such switching strategies depending on the target distribution have been investigated [9 , 13 , 14 , 42 , 50] , the response to predators is less understood [51] and may be a topic for further study . We introduced fitness determined not only by search efficiency but also by predation risk into the random search scenario unlike previous studies . In our assumption , the encounter with predators leads to death of the searcher with probability 1 . This means that the first encounter with the predator is crucial for the searcher , and seems to be more dangerous for the searcher than the actual situation in nature because the encounter in nature does not always lead to death . If the probability is less than 1 and the searcher survives the encounter with a predator , then the searcher starts to move from the position near the predator as the simplest assumption . In this case , the problem reduces to the difference of initial positions . The result of effects of initial distance between the searcher and the nearest predator suggests that the short distance decreases the relative encounter rate with predators γLWγBW when the predator’s strategy is sit-and-wait ( S6 Fig ) . Therefore , the Lévy walk strategy can temporarily benefit from departing from the close predator [30] , indicating that the switch between strategies could be more efficient . However , considering the biological plausibility , animals would not start to move around in a random manner immediately after an encounter with a predator . Instead , the searcher must depart from sit-and-wait predators using the information about the location of the predator in a deterministic manner , or dash to a safe area ( e . g . , bushes ) to hide from moving predators and wait for the predator to leave . The predators would leave the location after some giving-up-time . The encounter event with a predator seems to transcend the simple framework of the random search problem . However , if the searcher starts random searches after fully departing from predators , the condition should not change much . Thus , probability 1 can represent several situations of prey-predator interactions . Although we can use the probability of the survival for simplification , more complex interactions between prey and predator should occur in nature . Some empirical studies have reported the variability in predator avoidance [52 , 53] , and theoretical studies have solved the pursue-evasion problem [36 , 54] . Although the issue of how the random search problem relates to such complex interactions is an interesting one , the relationship is poorly understood at present , and awaits further study . Tracking animal movements over a prolonged period of time ( biologging ) is a method developed within the last decade that can lead to the understanding of dynamic phenomena ranging from the individual level to population and community levels [55 , 56] . Because the differences in searching strategies influence diffusiveness and movement patterns of animals , it is crucial to identify the search strategy that animals adopt in a natural environment . The tracking of animal movements within the framework of movement ecology requires information on biotic interactions and interactions between individual animals [57–59]; therefore , the context in our model should be common to various animal species in nature , because most animals are exposed to predation pressures or to the risk of death during searching . Likewise , predators may be exposed to the risks of higher-order predators . For further investigation , it will be interesting to explore the complex dynamics via the interactions between movement and population dynamics . Thus , considering ecological factors can lead to a fruitful understanding of the dynamics at various scales . | Moving agents should efficiently search for targets ( e . g . , food , prey , or specific locations ) when lacking information about the location of the targets . For this random search problem , the Lévy walk hypothesis claims that Lévy walk movement patterns ( i . e . , each step length follows a distribution that is heavy-tailed ) enable the searcher to capture more targets . However , most searchers may have antagonistic agents ( e . g . , predators ) that can lead to death . Thus , the searcher needs to seek targets while avoiding encounters with antagonistic agents . Here , we show that the Lévy search strategy is less efficient in terms of total lifetime fitness when the predators are abundant , and especially when predators adopt a sit-and-wait strategy . Moreover , the results indicate that the life-cycle type of the searcher is an important fitness factor . These ecological aspects significantly influence the consequences of the random search . Therefore , it is critical to consider the ecological properties of searchers and other interacting agents when examining and estimating animal movements . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Lévy Walks Suboptimal under Predation Risk |
Many pathogens , particularly those that require their host for survival , have devised mechanisms to subvert the host immune response in order to survive and replicate intracellularly . Legionella pneumophila , the causative agent of Legionnaires' disease , promotes intracellular growth by translocating proteins into its host cytosol through its type IV protein secretion machinery . At least 5 of the bacterial translocated effectors interfere with the function of host cell elongation factors , blocking translation and causing the induction of a unique host cell transcriptional profile . In addition , L . pneumophila also interferes with translation initiation , by preventing cap-dependent translation in host cells . We demonstrate here that protein translation inhibition by L . pneumophila leads to a frustrated host MAP kinase response , where genes involved in the pathway are transcribed but fail to be translated due to the bacterium-induced protein synthesis inhibition . Surprisingly , few pro-inflammatory cytokines , such as IL-1α and IL-1β , bypass this inhibition and get synthesized in the presence of Legionella effectors . We show that the selective synthesis of these genes requires MyD88 signaling and takes place in both infected cells that harbor bacteria and neighboring bystander cells . Our findings offer a perspective of how host cells are able to cope with pathogen-encoded activities that disrupt normal cellular process and initiate a successful inflammatory response .
The pathogen-associated molecular pattern ( PAMP ) hypothesis has been developed to explain how the innate immune system recognizes foreign microbial invaders . By this model , germline-encoded receptors recognize conserved foreign ligands associated with microbes , such as nucleic acids , lipopolysaccharide ( LPS ) , peptidoglycan or flagellin to generate a response directed at clearing the microorganism [1] , [2] . More recently , it has become clear that pattern recognition alone does not explain how multicellular organisms are able to differentiate virulent pathogens from harmless commensals and mount a response . It has been proposed that the host immune system can sense the presence of danger and respond to pathogen-encoded enzymatic activities that disrupt normal cellular processes . This mode of recognition , referred to as “effector triggered immunity” has been shown to play a significant role in pathogen clearance both in plants and mammalian cells [3] , [4] , [5] , [6] , [7] , [8] , [9] , [10] , [11] , [12] . Such recognition may be sufficient to activate a host response , but because it occurs simultaneously with PAMP recognition , host cell detection of pathogens likely results from integrating the recognition of microbial patterns together with pathogen-specific activities . Legionella pneumophila , the causative agent of Legionnaires' disease , promotes intracellular growth by translocating proteins into its host cytosol through its type IV ( Dot/Icm ) protein secretion machinery [13] , [14] , [15] . These translocated effectors serve various purposes , including recruitment of ER-derived membrane to the Legionella containing vacuole , inhibition of cell death pathways and manipulation of host lipid metabolism and regulatory pathways [16] , [17] , [18] , [19] , [20] , [21] . Most importantly for the innate immune response , after contact with macrophages , the bacterium stimulates a pathogen-specific response that is the consequence of simultaneous recognition of PAMPs and pathogen-translocated proteins that results in a unique response to this microorganism [10] . Legionella pneumophila is a pathogen for a broad range of fresh water amoebae , which provide the natural environmental niche for the microorganism and the source of exposure for humans [22] , [23] . After aspiration by a susceptible mammalian host , the bacterium is engulfed by alveolar macrophages in the lungs [24] . In cultured macrophages , L . pneumophila provokes signaling through various pattern-recognition receptors ( PRRs ) , such as Toll-like receptors ( TLRs ) and cytosolic NOD-like receptors ( NLRs ) [9] , [25] , [26] , [27] , [28] , [29] , [30] , [31] . This response is critical for clearance of the microorganism , because mouse mutants defective in these two responses succumb to lethal pneumonia [27] . Interestingly , macrophage challenge with wild type L . pneumophila ( Dot/Icm+ ) triggers a unique transcriptional response in host cells compared to mutants that lack a functional type IV secretion system , supporting the model that there is a pathogen-specific response involved in innate immune recognition [9] , [10] , [11] , [17] , [32] . Microarray studies have identified many of these transcriptional targets as being genes controlled by the NF-κB and mitogen-associated protein kinases ( MAPKs ) transcriptional regulators [9] , [17] , including downstream dual specificity phosphatases ( Dusp1 and Dusp2 ) , stress response genes ( Hsp70 , Gadd45a , Egr1 ) and pro-inflammatory cytokines and chemokines ( Il1α , Il1β , Tnfα , Il23a , Csf1 , Csf2 ) [9] , [10] , [11] , . It was recently demonstrated that the pathogen-specific response to Legionella is triggered by the action of L . pneumophila translocated effectors that interfere with host protein translation [10] , [11] . Disruption of the host translation machinery serves as a second signal ( in concert with signaling from PRRs ) to constitute the full innate immune response against Legionella pneumophila [10] . The elimination of five of these effectors is sufficient to block this response , even though it is clear that they are part of a much larger pool of translocated substrates that impinge on host protein synthesis [10] , [31] . These inhibitors ( the products of the lgt1 , lgt2 , lgt3 , sidI , and sidL genes ) modify eukaryotic elongation factor eEF1A and eEF1Bγ of mammalian cells and block protein synthesis both in vitro and in vivo [10] , [33] , [34] , [35] . In addition to blocking elongation , there is evidence that suggests wild type Legionella can also inhibit cap-dependent translation initiation [36] . Recognition of pathogenic Legionella leads to ubiquitination of the mTOR pathway , which in turn suppresses the eukaryotic initiation factor 4E ( eIF4E ) and prevents the synthesis of various genes [36] . This mode of translation inhibition was shown to induce translational biasing of host cells towards a more pro-inflammatory state [36] . However , it is currently not clear how host cells would be able to mount an inflammatory response when protein translation is blocked by L . pneumophila both at the initiation and elongation stages [9] , [17] . It is likely that pattern-recognition would play a role under conditions of intoxication , but the mechanism by which this is regulated is also unclear . A strong pro-inflammatory cytokine response is crucial for clearance of Legionella pneumophila [9] , [31] . The importance of cytokines can be seen in IL-1α , IL-12 , IFN-γ and TNF knockout mouse strains that show increased susceptibility to L . pneumophila infection [31] , [37] . Moreover , patients treated with TNF-α blockers are at high risk of developing severe Legionnaires' disease [38] . Given the key role that this innate immune response plays in clearance of L . pneumophila , we examined how cytokines and other immune mediators are synthesized under conditions in which the bacterium effectively blocks the host protein translation machinery . We find that although protein synthesis inhibitors induce the transcriptional response and block the translation of most genes , pro-inflammatory cytokine genes can bypass this blockade in a fashion that requires the TLR-adaptor protein MyD88 .
We hypothesized that Legionella-induced inflammatory gene transcription will be largely inconsequential , as the transcribed genes cannot be translated to proteins due to the bacterium-derived translation inhibitors . To test this hypothesis , we examined mammalian host protein synthesis predicted to be downstream of MAPK activation following exposure to L . pneumophila . Bone marrow-derived macrophages challenged with wild type L . pneumophila showed phosphorylation of MAPK members shortly after exposure to the bacterium , with the activation kinetics being almost identical to previous observations ( [9] , [39] , [40]; Figure S1 ) . In the first hour after L . pneumophila challenge , activation of JNK and P38 was independent of the Icm/Dot system , consistent with phosphorylation being the result of Tlr engagement [9] . A second wave of activation was observed beginning at two hours after challenge . This was dependent on the presence of the L . pneumophila type IV secretion system , as the levels of MAPK phosphorylation decayed when macrophages were challenged with the dotA3 mutant that lacks the Icm/Dot system ( Figure S1 ) . This two-wave activation , reflecting an early Tlr-dependent and a later L . pneumophila-specific response , mirrors our previous observations with NF-KB activation [41] . Cultured macrophages challenged with wild type L . pneumophila transcriptionally activate a number of dual specificity phosphatase ( DUSP ) genes , dependent on an intact Icm/Dot system [17] , [40] . Bone marrow-derived macrophages challenged with wild type Legionella for 4 hrs showed significant induction of the Dusp1 transcript compared to dotA3 infection ( Figure 1A ) . The transcriptional induction was much higher ( over 60 fold ) in U937 human monocytes , the cell lines in which dusp1 induction in response to wild type Legionella was first characterized ( Figure S2A ) [17] . This transcriptional response , however , was not accompanied by translation either in bone marrow-derived macrophages ( Figure 1B ) or U937 cells ( Figure S2B ) as DUSP-1 protein levels remained unchanged over the course of infection . Our inability to observe enhanced protein synthesis was not due to limitations with our detection system , because we observed a robust increase in DUSP-1 protein levels in response to the addition of LPS ( Figure 1C ) . Presumably the multiple translocated substrates that inhibit translation elongation [10] frustrate the transcriptional response , preventing translation of induced genes . To test this hypothesis , we challenged cells with an L . pneumophila mutant missing five of the translation inhibitors ( Δ5 ) to determine if the absence of these proteins could allow translation to proceed . Instead , we observed no transcriptional activation of the Dusp1 gene in response to this mutant ( Figure S2A ) . Therefore , induction of genes in the MAPK pathway and frustration of the response are tightly coupled . We then asked if the response leading to transcriptional upregulation of pro-inflammatory cytokine genes was similarly affected by translational inhibition . To understand how cytokines are regulated in response to L . pneumophila , we measured the transcription and translation of selected pro-inflammatory cytokines in C57/Bl6 ( B6 ) macrophages following L . pneumophila challenge . Flagellin deficient ( ΔflaA ) mutants were used in these experiments to avoid Caspase 1-dependent cell death downstream from NAIP5/NLRC4 recognition of flagellin by B6 macrophages [42] , [43] , [44] . Infection of bone-marrow macrophages with virulent ( Dot+ ) L . pneumophila induced Il1α , Il1β and Tnfα transcripts by 6 hrs post infection ( Figure 2A ) . As previously reported , the cytokine response to Dot+ was comprised of MyD88-dependent signaling that is layered on top of MyD88-independent , effector mediated signaling ( Figure 2A , black bars ) [10] . The response to the avirulent , dotA3 mutants on the other hand , was mostly dependent on MyD88 signaling ( Figure 2A , white bars ) [9] . To determine if the transcribed cytokine mRNAs were efficiently translated and secreted during infection , WT and MyD88−/− macrophages were challenged with L . pneumophila and cytokine protein levels were measured by western blot and ELISA . Contrary to what we saw for DUSP-1 , challenge with L . pneumophila Dot+ led to a significant increase in cell-associated pro-IL-1β levels after 4 hrs of infection ( Figure 2B ) . IL-1α and IL-1β mature forms could also be detected in culture supernatants after 24 hrs ( Figure 2C & 2D ) . Interestingly , challenge of macrophages with L . pneumophila dotA3 mutants accumulated pro-IL-1β transiently , with steady state levels reduced by 12 hr post infection ( Figure 2B densitometry ) , but this was not sufficient to induce the release of mature IL-1β . This is consistent with the hypothesis that in addition to TLR signaling , Icm/Dot translocated substrates are required for persistent pro-inflammatory cytokine activation and secretion [32] . Surprisingly , cytokine translation was severely diminished in MyD88−/− macrophages in response to wild type L . pneumophila ( Figure 2B ) despite the presence of large amounts of transcripts ( Figure 2A ) . Therefore , a MyD88-dependent signal appears necessary to bypass the translation block induced in response to wild type L . pneumophila . There is no clear model for why the presence of MyD88 allowed bypass of the translation block . Engagement of MyD88 on the host cell surface may lead to selective bypass of translation inhibition on a subset of transcripts . Alternatively , translation of cytokine transcripts could largely occur in neighboring uninfected cells that have not been directly injected with L . pneumophila translocated proteins , but which have been activated by bacterial fragments liberated by infected cells . We therefore asked if the observed cytokine translation was derived from neighboring bystander cells . B6 macrophages were challenged with L . pneumophila-GFP strains and macrophages harboring bacteria were sorted from uninfected bystanders . Cytokine transcripts ( Figure 3A ) and protein levels ( Figure 3B ) were measured in each population by qRT-PCR and Western blots . To ensure accumulation of TNF-α protein , cells were treated with GolgiPlug ( Brefeldin A; Materials and Methods ) to prevent secretion of this cytokine . No such treatment was necessary for IL-1α and IL-1β , which accumulate as precursors via an alternate secretion pathway [45] . Relative to cells that had never been exposed to bacteria , challenge of macrophages with L . pneumophila resulted in high levels of I1α , Il1β , Tnfα and Dusp2 transcripts in both infected ( GFP+ ) and neighboring uninfected ( GFP− ) populations by 4 hrs post infection , ( Figure 3A ) . Depending on the cytokine , the amount of transcription in the bystander cells varied from 10% −30% of that observed in the cells harboring bacteria . Dusp1 on the other hand , was mainly transcribed in GFP+ cells ( Figure 3A ) . More importantly , despite the presence of the translocated protein synthesis inhibitors , macrophages harboring bacteria were able to produce high levels of pro-IL-1α and pro-IL-1β ( Figure 3B , GFP+ cells ) . The kinetics of pro-IL-1α and pro-IL-1β production in infected cells indicated that there was enhanced accumulation of these proteins between 4–6 hrs post-infection ( Figure 3B , bottom panel ) . We will show that during this time window , L . pneumophila translation inhibitors effectively block most protein synthesis in infected cells ( below , Figure 4B ) . The presence of persistent cytokine synthesis in infected cells was confirmed by intracellular cytokine staining . B6 WT and MyD88−/− macrophages were challenged with L . pneumophila-GFP strains and intracellular cytokine levels were measured by flow cytometry . TNF-α was produced by both macrophages bearing bacteria ( GFP+ ) and bystander cells ( GFP− ) after challenge with L . pneumophila Δfla , while the dotAΔfla strain induced much lower levels of this cytokine ( Figure 3C , top panel 2nd and 3rd boxes ) . Cells harboring bacteria ( GFP+ ) were a significant source of IL-1α ( Figure 3C , bottom panel , 2nd box ) . Approximately 50% of the cells harboring bacteria showed detectable accumulation of IL-1α ( Figure 3C , bottom panels , 2nd box ) , while bacteria were associated with approximately 33% of the IL-1α-producing cells . Consistent with Figure 2B , translation of IL-1α and TNF-α were both dependent on MyD88 signaling , and accumulation of IL-1a in infected cells was dependent on the presence of the Icm/Dot translocator ( Figure 3C , two rightmost boxes in each panel ) . Time course analysis of intracellular IL-1α levels using flow cytometry confirmed that the highest level of IL-1α accumulation occurred between 2–6 hrs post infection in the GFP+ population ( Figure 3D and 3E ) . During this time period , the number of cells bearing bacteria that accumulated IL-1α increased from 1% of this population to approximately 50% ( Figure 3D ) . DUSP-1 protein levels on the other hand , remain unchanged between 2–6 hrs ( Figure 3E ) . The two signals that are received by mammalian cells during L . pneumophila infection ( 1st signal from TLR activation and 2nd signal from protein translation inhibition ) synergize to induce the full cytokine response against the bacterium . It was previously reported that pharmacological inhibitors of host protein translation induce transcription of various stress response genes and cytokines such as IL-6 , IL-23 , IL-α and IL-1β [10] , [31] . We wanted to confirm that translation and secretion of these cytokines could always bypass translation inhibition using the protein synthesis inhibitor cycloheximide ( CHX ) . CHX interferes with protein translation elongation by binding to the E-site of the 60S ribosomal subunit and preventing tRNA translocation [46] . Macrophages were treated with heat-killed Yersinia ( HKY ) to induce TLR signaling , together with 10 µg/mL of cycloheximide . Addition of the chemical inhibitor at the same time as HKY led to a complete inhibition of TNF-α and IL-1α production in bone-marrow macrophages ( Figure 4A ) . Contrary to what we observed during L . pneumophila infection , addition of CHX dampened the signal received from TLR stimulation ( Figure 4 A , B ) . Surprisingly , even at low concentrations of CHX that permit significant levels of protein translation ( Figure S3 ) , CHX was still able to inhibit IL-1β translation ( Fig . 4B ) . To rule out the possibility that the reduction in IL-1β levels during CHX treatment was due to cell death , we lowered the CHX dose to 0 . 5 µg/mL ( inhibits less than 50% of total host protein synthesis ) ( Figure S3 ) and also incubated the cells with the apoptotic inhibitor Z-VAD-FMK ( pan-caspase inhibitor ) . CHX was still able to inhibit IL-1β under these conditions , although it was clear that increased survival of cells was accompanied by higher accumulation of HKY-induced cytokine ( Figure 4C ) . MyD88-dependent stimulation of mouse macrophages in response to L . pneumophila primarily occurs via Toll-like receptor 2 ( TLR2 ) [29] . A recent report was able to reconstruct the cytokine induction seen during L . pneumophila infection by using the synthetic TLR-2 ligand Pam3CSK4 in combination with Exotoxin A ( Exo A ) , a toxin from Pseudomonas aeruginosa that interferes with translation elongation [31] . Based on this observation , we wanted to determine if specific activation of TLR2 is what leads to cytokine translation in the presence of protein synthesis inhibitors . Macrophages were treated with the TLR2 agonist Pam3CSK4 and pro-IL-1β levels were measured in the presence or absence of the protein synthesis inhibitor cycloheximide ( CHX ) . Drug treatment after addition of the TLR2 agonist led to a large reduction in pro-IL-1β levels ( Figure 4D ) . We also observed a failure to hyperstimulate pro-IL-1b in the presence of another protein elongation inhibitor , puromycin ( Figure S4 ) . Similar results were obtained when macrophages were stimulated with another TLR2 agonist lipoteichoic acid ( LTA ) , or a TLR4 agonist LPS , followed by addition of CHX ( data not shown ) . This indicates that the selective synthesis of cytokines may result from host cells sensing a specific mode of protein synthesis inhibition . It is also possible that the selective synthesis of pro-inflammatory cytokines is triggered by a block in translation initiation [36] instead of translation elongation , which would explain why the elongation inhibitors cycloheximide or puromycin were not able to induce the response . We considered two possible explanations for how cytokines were translated in the presence of the bacterium-derived protein synthesis inhibitors: ( 1 ) translation inhibition by L . pneumophila is not efficient , allowing most of the cytokine to be synthesized prior to a complete block; or ( 2 ) the host preferentially translates a subset of genes after protein synthesis is shutdown by pathogens . To distinguish between these possibilities , relative levels of protein synthesis were measured in bone-marrow macrophages at various times following L . pneumophila challenge , using an immunofluorescence readout . B6 macrophages were challenged with L . pneumophila flaA−-GFP at MOI = 10 for 2 hrs and translation of proteins was measured by incorporation of a methionine analog ( L-azidohomoalanine , AHA ) for an additional 4 hrs . Incorporated AHA was detected by reaction with a fluorescent-labeled phosphine reagent ( phosphine-APC ) , which covalently links to the azido-functional group on AHA ( Staudinger ligation reaction ) [47] . Protein translation was significantly inhibited in macrophages harboring L . pneumophila ( ΔflaA ) compared to bacteria lacking the Icm/Dot system ( Figure 5A ) . Furthermore , in macrophage cultures incubated with L . pneumophila , the macrophages harboring L . pneumophila showed selective protein synthesis interference , while the majority of the uninfected cells showed efficient incorporation of the amino acid analog ( Figure 5B; compare GFP+ to GFP− population ) . To determine the time point at which the protein synthesis inhibitors fully shut down global protein translation , pulse-chase experiments were performed in which the methionine analog ( AHA ) was added for 1 hr intervals starting at 2 hrs post infection ( Figure 5C ) . Between 2–3 hrs post-infection , approximately 40% of the cells harboring L . pneumophila were found in the population that has high levels of protein synthesis . Between 3–4 hrs post infection , we observed a major shift where almost 90% of cells harboring ΔflaA were found in the population having highly depressed protein synthesis . Later time points showed no further blockage in translation , perhaps reflecting the fact that there is a small fraction of wild type bacteria that fail to form replication compartments [48] . This population is predicted to show no significant translocation via the Icm/Dot system and should fail to inhibit protein synthesis ( seen in ∼10% of the macrophage population ) . To determine if translation-blocked cells could still produce cytokines , macrophage monolayers were challenged with L . pneumophila-GFP+ and protein synthesis was monitored by addition of AHA . The cells were then probed for IL-1α accumulation by immunofluorescence and flow analysis . In the infected GFP+ population of macrophages , the majority of cells that accumulated IL-1α show evidence of an almost complete shutdown of protein translation ( Figure 5E; red box ) . In the absence of MyD88 signaling , both the infected and uninfected populations showed little IL-1α accumulation ( Figure 5F ) . In contrast to IL-1α , there was no DUSP-1 accumulation after L . pneumophila challenge of macrophages ( Figure 5D ) . This confirmed our main hypothesis that there is translation of selected cytokine genes when protein synthesis is inhibited by L . pneumophila and this bypass requires MyD88 signaling . We have shown that pro-IL-1α and pro-IL-1β accumulate in cells that harbor L . pneumophila between 4–6 hrs post infection ( Figure 3B ) , despite a significant block in protein translation ( Figure 5C & E ) . To confirm that the accumulation we observe in infected cells was due to newly synthesized cytokines over the course of infection , we took advantage of another protein translation assay , SunSET . This assay uses puromycin incorporation into growing polypeptide chains to monitor active protein synthesis [49] . We modified this assay to measure the amount of puromycin incorporated into our protein of interest during a 1-hour pulse period . Accordingly , macrophages were challenged with L . pneumophila for increasing lengths of time , cells were labeled for one hour with 10 µg/mL of puromycin , lysed , and individual proteins were immobilized in assay wells using specific antibodies ( Materials and Methods ) . An ELISA was then used to determine the amount of puromycin incorporated in the immobilized proteins ( Figure 5G ) . Consistent with our intracellular cytokine staining and Western blot data ( Figure 3 ) , the highest levels of IL-1α and IL-1b synthesis were detected between 5–6 hrs post-infection . On the other hand , no significant puromycin incorporation was detected for DUSP-1 and RhoGDI proteins , confirming that there is selective synthesis of few genes after L . pneumophila challenge . To determine the role that L . pneumophila translation inhibitors play in modulating host cytokine synthesis , we used a mutant that lacks the 5 Icm/Dot translocated substrates known to block host protein synthesis ( Δ5 mutant ) [10] . The level of protein synthesis was first measured using AHA incorporation ( Figure 6A , B ) and puromycin incorporation ( Figure S5 ) in macrophages infected with Δ5ΔflaA-GFP+ . Compared to Dot+ strain ( Figure 6B & Figure S5 ) , there was an increase in active protein translation in cells that were infected with the Δ5 mutant , although the cells showed lower levels of protein synthesis than Dot− infected cells ( Figure 6B ) or the uninfected population ( Figure 6A; compare GFP+ to GFP− populations ) . This is consistent with the hypothesis that in addition to Icm/Dot translocated substrates that act on translation elongation , infection with virulent L . pneumophila also blocks translation initiation [36] . It had been previously reported that in the absence of most known pathways of pattern recognition ( MyD88−/− Nod1−/− Nod2−/− macrophages ) , the cytokine transcriptional response to L . pneumophila was primarily due to the presence of the translocated protein synthesis inhibitors [10] , [11] . Using macrophages that are only defective for MyD88 , this dependence on the translation inhibitors could be clearly observed for the Il1α , Il1β and Tnfα transcripts ( Figure 6C , MyD88−/− , gray vs . black bars ) . Consistent with our previous data , the transcriptional response in MyD88 knockout macrophages was unproductive , with no evidence that these highly induced transcripts are translated ( Figure 6E ) . In the case of wild type macrophages , Dot+ and Δ5 infections induced comparable levels of MyD88-dependent cytokine transcription and translation ( Figure 6C and E ) , and this could be observed in macrophages that were sorted by flow cytometry , as well ( Figure 6F ) . This result is in contrast with macrophages lacking MyD88 signaling , in which it is clear that there is protein synthesis inhibitor-dependent induction of cytokine transcripts ( Figure 6C ) , but this induction produces no apparent cytokine translation products . Interestingly , unlike what we see for dotA mutants , infection with Δ5 was still able to induce secretion of mature IL-1α at 24 hrs after infection [31] , even though it was to a lesser extent than wild type ( Figure 6D ) . Therefore , although the protein synthesis inhibitors are responsible for the transcriptional response that occurs in the absence of pattern recognition receptors , the production of cytokine proteins associated with infections by fully virulent strains is not dependent on these translocated substrates . Cytokine expression is regulated at various stages , including transcription , post-transcriptional processing , translation and secretion . One of the main regulatory steps for IL-1 and TNF production is their transcript stability , which is controlled by their AU-rich elements ( ARE ) in their 3′-noncoding regions and by various ARE-binding proteins [50] , [51] . To determine if lack of cytokine translation in MyD88−/− macrophages was due to mRNA instability , the half-life of Il1β and Tnfα transcripts were compared in WT and MyD88−/− macrophages after L . pneumophila challenge . Macrophages were first infected with L . pneumophila for 2 . 5 hrs , actinomycin D was added to the medium to block further transcription , and the amount of transcript remaining was measured by qRT-PCR at various time points after the addition of the drug . Tnfα mRNA was highly unstable in the absence of MyD88 when compared to MyD88+/+ macrophages ( Figure 7A ) . However , Il1β transcripts were relatively stable in the absence of MyD88 , and the amount of mRNA remaining after 2 hrs was similar to control macrophages ( Figure 7A ) . This indicates that the translation inhibition bypass of Il1β was independent of mRNA stability ( Figure 7 ) . In the case of tnfα , however , transcript stabilization via a MyD88-dependent signal may play a role in bypassing translation inhibition . Another possible explanation for why few pro-inflammatory cytokines , such as IL-1α and IL-1β , bypass translation inhibition could be mRNA abundance . A recent study has shown that these cytokines are the most abundant transcripts within macrophages after challenge with L . pneumophila [36] . Although there was significant transcription of these genes in MyD88−/− macrophages , the total mRNA abundance of IL-1α , IL-1β and TNF was still significantly lower compared to wild type macrophages ( Figure 2 ) . To address if mRNA abundance plays a role in bypass of translation inhibition during Legionella infection , WT and MyD88−/− macrophages were pre-activated with the TLR3 agonist poly ( I∶C ) for 2 hrs to induce NF-kB signaling . We also used heat-killed Yersinia pseudotuberculosis ( HKY ) to activate NF-kB via TLR4 . 2 hr pre-stimulation with 50 µg/mL poly ( I∶C ) was sufficient to trigger IL-1β transcription ( Figure 7B , left graph , white bars ) both in wild type and MyD88−/− macrophages . Pre-stimulation with poly ( I∶C ) followed by L . pneumophila infection increased Il1β transcription initially ( after 2 hrs of infection ) in wild type macrophages compared to cells that were untreated ( Figure 7B , compare black bars with grey bars ) . Surprisingly , this was reversed by 6 hrs post infection and cells that were pre-treated with poly ( I∶C ) down regulated their Il1β transcription ( Figure 7B , right graph ) . This was more pronounced in MyD88−/− macrophages ( Figure 7B , compare black and grey bars ) . We observed a very similar phenomenon when cells were pre-treated with heat-killed Yersinia followed by Legionella infection ( Figure 7D ) . As expected , pro-IL-1β translation was robust in wild type macrophages that were pre-treated with poly ( I∶C ) or with HKY followed by Legionella infection ( Figure 7C and 7E ) . Interestingly , pro-IL-1β was detected by Western blots in MyD88−/− macrophages after pre-stimulation with poly ( I∶C ) but the protein level was reduced when the pre-activated cells were challenged with L . pneumophila ( Figure 7C ) . MyD88−/− macrophages that were infected with Legionella alone on the other hand , showed little detectable translation ( Figure 7C ) despite having higher levels of Il1β transcripts ( Figure 7B , right graph , compare black and grey bars ) . This phenotype was more obvious during pre-treatment with HKY . 2 hr HKY pre-stimulation led to a robust pro-IL-1β translation initially , which was significantly reduced when cells were challenged with L . pneumophila ( Figure 7E ) . Therefore , in the absence of MyD88 signaling , macrophages were unable to overcome the translation inhibition induced by Legionella pneumophila even in the presence of external stimuli .
Inhibition of protein translation is a common virulence mechanism used by many viruses and bacteria . In this study , we showed that host cells have evolved mechanisms to cope with translation inhibition by selectively translating a subset of cytokine genes , including pro-inflammatory cytokines such as IL-1α and IL-1β in response to L . pneumophila challenge . The ability to bypass L . pneumophila translational inhibition is an important determinant of host protection , as mice defective in the IL-1α/IL-1β response and humans exposed to TNF-α inhibitors are highly susceptible to L . pneumophila infection [31] , [38] . L . pneumophila challenge of bone-marrow macrophages leads to a dramatic reduction in global protein translation ( Figure 5A ) . The bacterium interferes with protein translation both at the initiation step [36] and elongation step [10] . It has been shown previously that this inhibition triggers the transcription of various stress response genes including NF-kB- and MAPK-regulated genes , heat shock proteins and pro-inflammatory cytokines and chemokines [9] , [10] , [11] . We show here that L . pneumophila translocated effectors prevent the translation of these genes , resulting in a “frustrated response , ” in which there is accumulation of transcripts but no increase in protein levels . A subset of cytokine genes , and potentially other genes that have not been identified yet , are insensitive to this inhibition and get translated in cells that show the highest level of protein synthesis inhibition . We observed bypass of translational inhibition as an orderly series of events resulting in the accumulation of IL-1α in cells harboring bacteria as well as in bystanders , followed by release of the cytokine into culture supernatants . Initial transcriptional induction and translation of IL-1a occurred independently of the Icm/Dot system , and was associated with TLR-signaling , consistent with the TLR-dependent activation of the NF-κB response known to occur at early time points after L . pneumophila challenge [41] . This was followed by persistent accumulation of pro-IL-1α protein in a process that required both the presence of the Icm/Dot system and MyD88-dependent signaling , indicating collaborative signaling between the two pathways . Surprisingly , accumulation of pro-IL-1α was equally robust in both cells harboring bacteria and in bystander cells , in spite of the translocated protein synthesis inhibitors that are deposited by L . pneumophila . This observation is particularly striking , in that time points showing the strongest inhibition of protein synthesis also resulted in the fastest rate of pro-IL-1α accumulation , arguing that there is selective ribosomal loading of cytokine transcripts in intoxicated cells . In the absence of MyD88 signaling , no such bypass could be observed in either infected or uninfected cells , either because translation requires the extremely high levels of transcription that occur in the presence of MyD88 , or there is pattern recognition-dependent bypass of translational inhibition . Accumulation of pro-IL-1α was then followed by its release , which required both the Icm/Dot system and MyD88 . It is conceivable that the L . pneumophila translational inhibitors could be responsible for the induction of pro-IL-1α . Arguing against this model is the fact that a strain that lacks 5 of the known translation inhibitors ( Δ5 ) still induced considerable pro-IL-α accumulation ( Figure 6E , F ) [31] . The accumulation of cytokine in response to this strain could have resulted from residual translation inhibition that was observed , but it should be noted that MAPK and NF-κB activation resulting from macrophage challenge by this strain is due to a pattern recognition response and is not due to Icm/Dot signals [11] . The initial MyD88-dependent activation that occurs after contact with L . pneumophila may be amplified by unknown Icm/Dot-signals or due to inhibition of translation initiation [36] . A recent study proposed that infection of macrophages with virulent L . pneumophila strains ( both Dot+ and Δ5 ) leads to downregulation of mTOR activity , which is sufficient to suppress cap-dependent protein translation initiation [36] . The second signal that is required for amplifying pattern-recognition could be generated from such translational suppression , and could be the trigger for induction of pro-IL-1α . Challenge of macrophages that lack MyD88 with L . pneumophila induces cytokine gene transcription , but the transcribed genes fail to be translated . In the absence of MyD88 , therefore , the protein translation inhibition takes on global dimensions . The MyD88-dependent bypass of the translation inhibition was independent of transcript stability in the case of Il1β transcript , which is a known strategy for post-transcriptional regulation of cytokines [52] , [53] . This surprising result indicates that there may be a previously unrecognized MyD88-dependent signaling pathway that mediates post-transcriptional regulation of cytokine transcripts . It has previously been reported that separate transcriptional and translational signals are required for IL-1β expression [54] . Although it is not clear what these translational signals could be , it is possible that MyD88-dependent signals could result in either enhanced ribosome loading , or could regulate translation via action at the 3′ or 5′ untranslated regions . Alternatively , the role of MyD88 could be totally passive , and merely a consequence of enhancing expression of cytokine gene transcripts . Although L . pneumophila infection causes a large induction of cytokine transcription in the absence of MyD88 , these levels are still lower than what is seen when pattern recognition is intact ( Figure 2A ) . This added boost in cytokine gene transcription by MyD88 may be sufficient to push the concentrations of these transcripts above the minimum threshold necessary to support selective translation of these genes under conditions of intoxication . Consistent with this model are previous results from nanostring analysis of macrophage transcripts that are induced in response to L . pneumophila challenge [36] . In this work , it is argued that the primary determinant of translation in cells challenged with L . pneumophila is the relative abundance of a particular transcript . Translation was most likely to occur from transcripts that were the most abundantly expressed after bacterial challenge [36] . A similar phenomenon to that reported here has been observed in the model organism C . elegans upon infection with Pseudomonas aeruginosa . C . elegans intestinal cells endocytose P . aeruginosa Exotoxin A , which shuts down protein translation by inhibiting elongation factor 2 ( EF2 ) [8] . This inhibition leads to the selective translation of ZIP-2 , which is required for activation of defense pathways and pathogen clearance [7] . It was proposed that the 5′ UTR of zip-2 , which contains several untranslated ORFs ( uORFs ) , was required for the selective translation . Even so , there is no explanation for how ribosome loading and translation can selectively occur in this transcript . Interestingly , there are a few other examples from mammalian cells and yeast , where protein translation inhibition leads to selective translation of few genes that have uORFs at their 5′ UTR . The mammalian stress response transcription factor ATF4 and the yeast transcription factor GCN4 respond similarly to amino acid starvation and protein synthesis inhibition [55] , [56] . The 5′ UTR or 3′ UTR of cytokines could potentially be functioning the same way to allow selective protein translation when initiation and/or elongation is blocked by Legionella pneumophila . Interestingly , pharmacological inhibitors of host protein translation induce transcription of various stress response genes , including pro-inflammatory cytokines such as Il6 , Il23 , Il1α and Il1β [10] , [11] , [31] . Secretion of these genes can take place when the highly conserved host elongation machinery is targeted by toxins such as P . aeruginosa Exotoxin A or Corynebacterium diphtheriae encoded diptheria toxins [10] , [31] . Diphtheria toxin and Pseudomonas ExoA inhibit eukaryotic elongation factor similar to the mechanism used by L . pneumophila effectors . They modify elongation factor 2 ( EF2 ) of eukaryotic cells by ADP-ribosylation , which has been shown to trigger a strong host immune response [31] , [57] . This suggests the presence of a conserved surveillance mechanism the host uses to detect and respond to inhibition of the translation elongation machinery . A previous study has shown that the cytokine induction seen during L . pneumophila can be mimicked by the addition of P . aeruginosa Exo A in combination with the synthetic PAMP ligand Pam3Csk4 [31] . It seems likely that macrophages are able to selectively bypass the translation block of this toxin in a fashion that is similar to that described in C . elegans . Interestingly , we cannot reproduce this result using a variety of concentrations of the protein synthesis inhibitor cycloheximide , which instead reduces pro-inflammatory cytokine production in response to PAMP challenge . The mechanism by which cycloheximide inhibits protein synthesis is sufficiently different from these toxins to explain why we see differences in the innate immune response against CHX . CHX binds to the E-site of the 60S ribosomal subunit and freezes all translating ribosomes [46] . The RNA/ribosome complex remains stabilized and does not dissociate , a phenomenon that may not be perceived as danger by eukaryotic cells . In contrast , both L . pneumophila and ExoA interfere with elongation factor function . There may be a set of modified elongation factors in the host cell that resist action of these effectors , or there may be a population that is sequestered from modification , allowing them to selectively act on cytokine transcripts . In either case , there must be some special property to the cytokine transcript that allows this selective utilization of these active elongation factors . Future work will focus on the nature of these transcripts that allows bypass of translation inhibition .
This study was carried out in accordance with the recommendation in the Guide for Care and Use of Laboratory Animals of the National Institutes of Health . The Institutional Animal Care and Use Committee of Tufts University approved all animal procedures . Our approved protocol number is B2013-18 . The animal work , which is limited to the isolation of macrophages , does not involve any procedures of infections of live animals . L . pneumophila strains Lp02 ( referred to as WT ) and Lp03 ( referred to as dotA3 ) are streptomycin-resistant restriction-defective thymidine auxotrophs derived from L . pneumophila Philadelphia-1 ( Lp01 ) ( Table 1; [58] ) . The Δ5 and Δ5 . ΔflaA strains were kindly provided by Zhao-Qing Luo ( Purdue University ) [10] , [11] , [43] . ΔflaA-GFP+ , dotAΔflaA-GFP+ , Δ5 . ΔflaA-GFP+ carry GFP on an isopropyl-β-D-thiogalactopyranoside ( IPTG ) –inducible , Cm resistant plasmid ( Table 1; [17] , [59] ) . Solid medium containing buffered charcoal yeast extract ( BCYE ) and ACES-buffered yeast extract ( AYE ) broth culture medium supplemented with 100 µg/mL thymidine were used to maintain L . pneumophila strains [60] , [61] . Strains containing the pGFP plasmid were maintained on BCYE plates containing 100 µg/mL thymidine and 5 µg/mL chloramphenicol and grown in AYE containing 100 µg/mL thymidine , 5 µg/mL chloramphenicol and 1 mM IPTG [17] . Bone marrow-derived macrophages ( BMDMs ) were isolated from the femurs of mice and allowed to proliferate as described [60] , [61] . C57BL/6 myd88−/− femurs were kindly provided by Tanja Petnicki-Ocwieja in the laboratory of Linden Hu ( Tufts Medical Center ) . BMDMs were differentiated for 7 days in RPMI containing 30% L-cell supernatant , 10% FBS , 2 mM L-glutamine and 1× Pen Strep ( 100 U/mL penicillin , 100 µg/mL streptomycin ) . Cells were lifted and either re-plated for experiments or quick-frozen for later use in FBS and 10% DMSO . U937 cells ( ATCC ) were grown in RPMI supplemented with 10% FBS and 1 mM L- glutamine . For differentiation , cells were treated with 10 ng/ml 12-tetradecanoyl phorbol 13-acetate ( TPA ) for 48 hrs . For L . pneumophila infections , U937 cells were plated in fresh media without TPA and infections were carried out 12–16 hours after plating . To evaluate protein expression in host cells , C57BL/6 bone marrow-derived macrophages were plated in medium supplemented with 200 µg/mL of thymidine . Cells were challenged with L . pneumophila at the desired MOI , subjected to centrifugation at 1000×g for 5 min and incubated at 37° for the noted time periods . Lysates were collected using 2× SDS Laemmli sample buffer ( 0 . 125 M Tric-Cl pH 6 . 8 , 4% SDS , 20% glycerol , 10% beta-mercaptoethanol , 0 . 01% bromophenol blue ) . Proteins were electroblotted to PVDF membranes , blocked in milk and analyzed by immunoprobing . For phospho-specific antibodies , blots were washed of all milk and incubated overnight with 1∶1000 phospho-p38 or phospho-JNK ( Cell Signaling ) in 5% BSA in phosphate buffered saline ( PBS ) . Rabbit anti-DUSP1 ( MKP1 V-15 , Santa Cruz ) and mouse anti-tubulin ( sigma ) antibodies were diluted to 1∶200 and 1∶7 , 500 , respectively , in 5% milk in TBST . Goat anti-IL-1α and Goat anti-IL-1β antibodies ( R&D systems , AF-400-NA & AF-401-NA ) were diluted to 1∶500 in 5% milk in TBST . For preparation of Heat-Killed Yersinia , wild type Y . pseudotuberculosis strains were grown overnight at 26° in Luria-Bertani ( LB ) broth . Overnight cultures were heat killed at 60° for 30–60 min and aliquots were frozen at −80° until use . Macrophages were stimulated with HKY ( MOI = 50 ) , LPS ( Sigma , 0 . 1 µg/mL or 1 µg/mL ) and Pam3CSK4 ( Invitrogen , 2 µg/mL ) , poly ( I∶C ) ( InvivoGen , 50 µg/mL ) for the desired time points . Cells were washed 3× with PBS and lysed in 2× SDS Laemmli sample buffer . RNA was extracted from mammalian cells using RNAeasy kit ( Qiagen ) . To determine the amount of a particular transcript , a one step , RNA-to-Ct kit ( Applied Biosystems ) was used according to manufacturer's instructions . Primers used for transcript analysis were as follows: human Dusp1 ( 5′ TTTGAGGGTCACTACCAG and 3′ CCGCTTCGTAGTAGAG ) , mouse Dusp1 ( 5′GGATATGAAGCGTTTTCGGCT and 3′ ACGGACTGTCACGTCTTAGG ) , mouse Il1α ( 5′GCACCTTACACCTACCAGAGT and 3′ TGCAGGTCATTTAACCAAGTGG ) , mouse Il1β ( 5′ GCAACTGTTCCTGAACTCAACT and 3′ ATCTTTTGGGGTCCGTCAACT ) , mouse Tnf-α ( 5′ GCACCACCATCAAGGACTCAA and 3′ GCTTAAGTGACCTCGGAGCT ) , mouse 18S ribosomal RNA ( 5′ CGCCGCTAGAGGTGAAATTCT and 3′ GCTTTCGTAAACGGTTCTTCA ) . Macrophages were plated in 24 well tissue culture plates ( 2 . 5×105 per well ) and challenged with L . pneumophila for either 6 hrs or 24 hrs . Supernatants were collected from each sample and 50 µL was used for ELISA . Mouse IL-1α and IL-1β Platinum ELISA ( eBiosciences ) was used to measure cytokine levels according to the manufacturers manual . Intracellular cytokine staining was performed as described before [62] with modifications . Differentiated macrophages ( ∼1×107cells/plate ) were washed and challenged with L . pneumophila GFP+ strains in RPMI containing 200 µg/mL thymidine , 5 µg/mL Chloramphenicol and 1 mM IPTG . For measuring TNF production , infections were carried out at MOI-3 and MOI-10 for 9 hrs followed by Golgiplug incubation ( BD Bioscience , 1 µL/mL ) for additional 5 hrs . For IL-1α and IL-1β , infections were carried out for 6 hrs . Cells were harvested with 10 mL cold PBS , washed twice with FACS buffer ( PBS+0 . 5%BSA+0 . 05%NaN ) , incubated with Fc Block ( clone 2 . 4G2 ) for 20 minutes and fixed with 2% paraformaldehyde overnight at 4° . Macrophage were permeabilized with Perm/Wash buffer ( BD Bioscience ) on ice for 20 minutes and stained with Alexa Fluor 647-conjugated anti-mouse TNF ( BioLegend , clone ALF 161 ) , phycoerythrin ( PE ) -conjugated anti-mouse IL-1α ( BioLegend , clone MP6-XT22 ) , PE-conjugated anti-mouse IL-1β ( eBioscience , clone NJTEN3 ) , Rabbit anti-DUSP1 ( MKP1 V-15 , Santa Cruz ) and Goat anti-Rabbit Cy5 ( Invitrogen ) for 40 minutes . Stained cells were analyzed by BD LSR II flow cytometer . To detect whole cell translation during defined timepoints , C57BL/6 bone marrow-derived macrophages were challenged with L . pneumophila-GFP+ at MOI-10 for 2 hrs , followed by labeling cells at various timepoints with 50 µM of L-azidohomoalanine ( AHA ) ( Invitrogen ) added to the culture medium . Cells were incubated for 1 hr ( time course experiments ) or for 4 hrs ( IL-1α co-staining experiments ) to allow incorporation of AHA into growing polypeptide chains . At the end of each incubation period , cells were washed with PBS , fixed with 4% paraformaldehyde for 15–20 min and left overnight in PBS . Incorporation of AHA was monitored by a Biotin- or APC-conjugated phosphine reagent ( Pierce ) and nuclei were stained with Hoechst 33342 ( Molecular Probes ) . For flow cytometry , fixed cells were blocked with 1× BSA/PBS for 30 min at RT and 100 µM of APC-phosphine was added . Cells were incubated at 37° for 2–3 hrs and excess dye was removed by washing with 0 . 5% Tween-20/PBS . For IL-1α co-staining , PE conjugated anti-mouse IL-1α ( BioLegend , clone MP6-XT22 ) was added to cells on ice for 30 min . Cells were analyzed by BD LSR II flow cytometer or by BD FACScalibur . SunSET assay was used to determine the kinetics of protein translation over time [49] with modifications . Macrophages were plated in 6 well plates and infected with ΔflaA for the desired time points . 10 µg/mL of puromycin ( Sigma ) was added to cells for either 15 min or 1 hr . Macrophages were washed and lysed with IP Wash/Lysis buffer ( Pierce ) in the presence of protease inhibitors ( Roche ) . Lysates were incubated on ice for 20 min , centrifuged at 13 , 000 rpm for 5 min and the supernatants were used for ELISA . The protein concentration in each sample was determined by Bradford assay . ELISA plates were prepared by coating 96 well Nunc MaxiSorp plates with the desired antibody . Polyclonal Goat anti-IL-1α and IL-1β ( R&D systems ) , Rabbit anti-DUSP-1 ( Santa Cruz ) and Rabbit anti-RhoGDI ( Santa Cruz ) antibodies were diluted in Carbonate/Bicarbonate buffer ( PH = 9 . 6 ) to 10 µg/mL and 100 µL was used per well . Plates were incubated overnight at 4° . The following day , plates were brought to room temperature and blocked with 0 . 5% BSA/PBS for 1 hr . Cell lysates that have incorporated puromycin were incubated on the ELISA plates for 2 hrs , washed with 0 . 05% Tween-20/PBS three times and incubated with monoclonal mouse-anti-Puromycin ( 12D10 , Millipore ) for 1 hr . Unbound antibody was washed exhaustively with 0 . 05% Tween-20/PBS and plates were incubated with Donkey anti-mouse-HRP secondary antibody . Unbound antibody was washed again with Tween-20/PBS four times and 100 µL of HRP substrate ( TMB solution ) was added to each well for 5–10 mins . The reaction was stopped by adding 100 µL of stop-solution ( 2N H2SO4 ) , and absorbance was measured at 450 nm . | Translation inhibition is a common virulence mechanism used by a number of pathogens ( e . g . Diphtheria Toxin , Shiga Toxin and Pseudomonas Exotoxin A ) . It has been a mystery how host cells mount a pathogen-specific response and clear infection under conditions where protein synthesis is blocked by pathogens . Using Legionella pneumophila as a model , a bacterium that efficiently blocks the host protein translation machinery , we show here that the innate immune system has devised a mechanism to cope with translation inhibition by selectively synthesizing proteins that are required for inflammation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"cell",
"biology",
"genetics",
"biology",
"and",
"life",
"sciences",
"immunology",
"microbiology",
"molecular",
"cell",
"biology"
] | 2014 | The Frustrated Host Response to Legionella pneumophila Is Bypassed by MyD88-Dependent Translation of Pro-inflammatory Cytokines |
About 85% of the maize genome consists of highly repetitive sequences that are interspersed by low-copy , gene-coding sequences . The maize community has dealt with this genomic complexity by the construction of an integrated genetic and physical map ( iMap ) , but this resource alone was not sufficient for ensuring the quality of the current sequence build . For this purpose , we constructed a genome-wide , high-resolution optical map of the maize inbred line B73 genome containing >91 , 000 restriction sites ( averaging 1 site/∼23 kb ) accrued from mapping genomic DNA molecules . Our optical map comprises 66 contigs , averaging 31 . 88 Mb in size and spanning 91 . 5% ( 2 , 103 . 93 Mb/∼2 , 300 Mb ) of the maize genome . A new algorithm was created that considered both optical map and unfinished BAC sequence data for placing 60/66 ( 2 , 032 . 42 Mb ) optical map contigs onto the maize iMap . The alignment of optical maps against numerous data sources yielded comprehensive results that proved revealing and productive . For example , gaps were uncovered and characterized within the iMap , the FPC ( fingerprinted contigs ) map , and the chromosome-wide pseudomolecules . Such alignments also suggested amended placements of FPC contigs on the maize genetic map and proactively guided the assembly of chromosome-wide pseudomolecules , especially within complex genomic regions . Lastly , we think that the full integration of B73 optical maps with the maize iMap would greatly facilitate maize sequence finishing efforts that would make it a valuable reference for comparative studies among cereals , or other maize inbred lines and cultivars .
Maize ( Zea mays ssp . mays L . ) is a pervasive , economically valuable crop supplying the world with food , animal feed , and with biofeedstocks used in the synthesis of a broad range of industrial products . It is also a model system for classical genetics and cytogenetics that has significantly contributed to our understanding of fundamental processes that include reproduction , photosynthesis , biosynthesis of primary metabolites , mobile elements , and chromosome structure-function relationships . Investigators have developed extensive genetic tools over the last two decades dealing with male sterility , QTLs , regeneration of crop species , wide hybridization , marker assisted selection , associative mapping , endosperm mutants , transgenic crops , genetic control of meiosis , transposable elements , chromosome elimination , etc . In addition , diverse germplasms have been accumulated that have leveraged the assessments of genomic modifications during domestication , molecular mechanisms of heterosis , and the roles played by mobile DNA elements affecting genome evolution . Such advances are now being rapidly exploited with paradigm shifting tools and resources that are fostering insights emerging from fully sequenced and annotated genomes . In 2005 three funding agencies – NSF , DOE and USDA – jointly pledged $32 million towards a 4-year program to sequence the maize genome . These agencies' goals were to ensure that cutting-edge genomic resources would be available for maize to accelerate translational research in the agriculture and bioenergy sectors . The maize genome is estimated to be 2 . 3–2 . 5 gigabases ( Gb ) in size [1] , and its architecture presents significant challenges for comprehensive sequencing . An intriguing attribute of the maize genome is its allotetraploidy nature that originated at least 5 million years ago ( mya ) from two progenitors , which had previously diverged from a common ancestor about 12 mya [2]–[3] . The maize genome underwent a whole genome duplication event in the hybridization of the two progenitors , and then gradually became diploid through loss of ∼50% of one of its progenitors' gene copies [4]–[8] . The architecture of the maize genome is also heavily punctuated by a complex motif of repetitive elements . About 85% of the genome is made up of a complex mix of repetitive DNA that mainly includes numerous families of retrotransposons such as Tekay , Huck , PREM-2 , Opie , Ji , etc . [9]–[11] . These retroelements mostly appeared during the last 1 to 3 million years and thus show great similarity [9] . The chromosome “knobs” consist of megabase-sized satellite sequences interspersed with retrotransposons , while the euchromatic regions harbor repetitive insertions of transposons , with most retrotransposons tending to insert within each other , resulting in nested retrotransposons in the intergenic regions [9] , [11]–[17] . Therefore , maize genes are like small islands surrounded by seas of nested retrotransposons , and such challenging attributes have necessitated development of multiple sequencing approaches . Given the current need for a broadly informative representation that includes coding sequences and precise physical characterization of gaps between genes , accurate genetic and physical maps are required for guiding the large-scale sequencing of maize genome . The genetic , physical , and integrated maps available for maize are briefly described . The 1935 maize genetic map featured just 62 loci that relied on morphological variants [18] . Advancements in new technologies and genomic insights led to the addition of nearly 6 , 000 markers to create a high resolution genetic map using the intermated B73 X Mo17 ( IBM ) populations [19]–[20] ( http://maize-mapping . plantgenomics . iastate . edu/ ) . This augmentation drew new resources from the development of cytological markers based on B-A translocations , molecular markers based on isozymes , restriction fragment length polymorphisms ( RFLPs ) , microsatellite or simple sequence repeat ( SSR ) markers , single nucleotide polymorphisms ( SNPs ) , and cDNA or expressed sequence tags ( EST ) markers [19] , [21]–[38] . In addition , FPC ( fingerprinted contigs ) [39] map contigs were also anchored to this genetic linkage map , and this highly integrated resource became known as the “iMap . ” Early physical mapping of maize used a YAC ( yeast artificial chromosome ) library constructed from an inbred line UE95 [34] . The YAC libraries proved to be of limited utility due to a significant level of clone chimerism , or issues surrounding YAC stability and faithful representation of genome copy number [40] . With the advent of stable large-insert cloning in bacteria , more reliable BAC ( bacterial artificial chromosome ) [41] libraries were constructed for the maize B73 inbred line . Clones were fingerprinted , hybridized with known molecular markers ( genetic ) , and FPC mapped . These efforts integrated the genetic and physical maps through assignment of molecular markers from the genetic map to individual BAC clones within FPC contigs [35] , [42]–[46] . FPC maps were later greatly refined by HICF ( fluorescent-based high-information-content fingerprinting ) mapping of these libraries which reduced the number of FPC contigs from 4 , 518 to 1 , 500 [47] . The number of FPC contigs was further reduced to 721 ( 2 , 150 Mb; May , 2006 ) by manual curation based on agarose-based fingerprinting and on knowledge gleaned from the HICF map and syntenic markers between the maize and rice genomes [37] . In addition to the ∼6 , 000 genetic markers , there are over 24 , 000 sequence markers integrated into the maize genetic-physical ( FPC ) map ( also termed iMap ) , including expressed sequence tag ( EST ) -derived unigene markers , overgos derived from maize EST sequences , conserved genomic sequences , and end-sequence data from gene-containing BACs . The inclusion of these sequence markers into the integrated map ( iMap ) ( IBM2; iMap; http://www . maizemap . org/iMapDB/iMap . html ) has greatly increased the marker density across the entire maize genome and created a framework for directed clone-based sequencing and assembly of chromosome-wide pseudomolecules [37] , [44] , [48] . However , the maize genome is structurally highly polymorphic , as seen in the significant structural variation among different inbred lines and even between different haplotypes [49]–[52] . Because the maize iMap integrates the IBM genetic map with the B73 inbred line FPC physical map , structural differences between the IBM population and the B73 genome ( targeted genome for sequencing http://ftp . maizesequence . org/release-3b . 50/All Releases/ [53] ) would be expected . The primary construction of high-resolution physical maps that are not dependent on the IBM genetic map would offer an essential resource for the comprehensive and accurate assembly of the maize B73 reference genome . Sequencing efforts for the maize genome have progressed through three stages: the pilot , gene enrichment , and clone-by-clone full genome sequencing stages . 1 ) The pilot sequencing effort considered large parts of chromosome arms and BAC-end sequence gathered from random clones; this provided an early glimpse into genome structure , organization , and sequence composition [48] , [54]–[55] . 2 ) The gene-enrichment approaches culled gene-rich templates for side-stepping notoriously complex sequence repeats and high copy number DNA elements present in the maize genome . Enrichment was accomplished by a variety of sequencing approaches that included ESTs , genome filtration ( methylation filtration and high-Cot selection ) , RescueMu ( RM ) , and hypomethylated partial restriction ( HMPR ) [56]–[63] . Sequence data enriched for genes , collectively termed as Genome Survey Sequences ( GSSs ) , are scattered throughout the maize genome , typically comprising small sequence contigs a few kilobases in size [61] . 3 ) In contrast , clone-by-clone sequencing used a comprehensive , hierarchical , map-based approach that allowed construction of a BAC minimal tiling path across the iMap . Tiled BACs were then individually shotgun-sequenced and assembled [12] , [64]–[66] . Although this map-based approach simplified assembly , an individual BAC assembly typically contained multiple unordered sequence contigs . Sequencing of the maize genome is now in the finishing phase with more than 16 , 000 sequenced BACs [53] . But complete sequencing and creation of a highly accurate assembly of the maize genome still hold daunting challenges for the maize community . A direct and encompassing way to deal with the formidable architecture of the maize genome is to analyze “chunks” of it , at high-resolution , that are as large as possible . In this way , nests of sequence repeats are largely bridged by chunks that offer a sufficient level of unique sequence information for supporting de novo genome assembly . With this concept in mind , we constructed a high-resolution optical map [67]–[76] that spans ∼91% of the maize genome by the de novo assembly of a large data set containing ordered restriction maps of individual genomic DNA molecules ∼500 kb in size . This ordered restriction map provides an independent resource that lays out an accurate physical metric across the entire maize genome . Because large molecules were analyzed , we were able to physically map repeat-rich regions and link sequence and map data within complex genomic regions . We show here that our maize optical map identifies gaps within and between sequence contigs and guides the assembly and validation of reference chromosomes .
We constructed a whole-genome shotgun optical map for maize using the CpG methylation insensitive restriction enzyme SwaI . The optical map data set contains 2 , 116 , 074 genomic DNA molecules , ranging in size from 300 kb to 3 , 700 kb , and totaling ∼927 , 604 Mb , or ∼403× coverage of the maize genome . The maps in this raw data set—one optical restriction map per genomic DNA molecule—have a mean length of 438 . 4 kb with an average fragment size of 26 . 1 kb . Because of the vast size of the maize optical data set , our de novo assembly of maps relied on a divide and conquer strategy that leveraged available cluster computing resources [77] . Briefly , we divided the raw map data set into 40 separate bins . Each bin was assembled into contigs and processed to remove redundant contigs and/or overlapping contigs , producing seed maps ( consensus maps ) for our iterative assembly scheme ( Materials and Methods ) . After five initial cycles of iterative assembly , the terminal 40 restriction fragments of a seed map ( Materials and Methods ) were selected for augmentation of optical contigs that were >10 Mb . These optical consensus maps were lengthened and their depth of coverage was increased through an additional 15 cycles of iterative assembly using the entire map data set . In this way , we constructed 66 optical consensus maps spanning a total of 2 , 103 . 93 Mb . The consensus maps were internally validated in an additional iterative assembly step . They were partitioned into a series of overlapping 1 Mb map intervals for use as new seed maps , with the overlaps covering ∼500 kb . Because this diagnostic assembly reproduced the original set of 66 parental contigs , the current optical assembly is apparently free of any chimeric maps . Statistics describing the 66 optical map contigs are shown in Table 1 . In total , 339 , 280 of the 2 , 116 , 074 maps were assembled into 66 optical map contigs . The average depth of coverage is 72 restriction fragments per contig ( Table 1 ) . The breadth of these contigs range from 3 . 64 Mb–100 . 76 Mb , and the average contig size is 31 . 88 Mb . The average size of restriction fragments of each contig ranges from 21 . 32 kb to 28 . 53 kb , with the overall size averaging 23 . 56 kb ( Table 1 ) . Lastly , the rate of contig formation was 16 . 03% , and we attribute this modest value to the modest rate of restriction digestion caused by unknown inhibitors within our DNA preps ( genomic; ∼500 kb sized molecules ) that attenuated restriction enzyme action . We leveraged the assembly process for overcoming this problem by increasing the number of digested molecules within the raw data set for biasing those molecules with adequate restriction patterns supporting confident contig formation . Optical contigs terminate to form a gap when the SwaI restriction site density is low , or when a contig reaches the end of a chromosome . Sharply demarcated contig edges may represent telomere associated sequences near chromosome ends . Using these criteria we identified 15 contigs ( OMcontigs_7 , 8 , 13 , 16 , 20 , 21 , 23 , 28 , 31 , 35 , 38 , 39 , 47 , 51 , and 61 ) that have reached the ends of chromosome as evidenced by contig “edges” comprising more than 5 maps that show no significant map “overhangs” ( Figure 1 ) . A collection of DNA molecules ( maps ) is said to overhang at a contig's end when their terminal restriction fragments are large and vary in length—such patterns describe gaps . The maize genome optical map contains 66 optical contigs and 91 , 453 ordered SwaI restriction fragments . However , placement of tiled ( 16 , 848 FPC clones ) , but unfinished , BAC sequences released by the maize genome sequencing project ( release 3b . 50; http://ftp . maizesequence . org/release-3b . 50/All%20Releases/; March 19 , 2009 ) on optical maps required development of a new algorithm that considers alignments of FPC clones comprising unordered and unoriented sequence contigs ( averaging 11 sequence contigs per BAC ) . We had developed a new algorithm several years ago to integrate the optical and FPC maps through the alignment of unfinished BAC sequence data ( Materials and Methods , Figure 2 ) . Our motivation for its development was to anchor large optical contigs to the iMap , which in 2007 contained only ∼6 , 000 sequenced BACs . The algorithm—named “BACop” —considers “complete” SwaI restriction fragments ( fragments having pairs of SwaI sites ) present in the in silico digest of BAC sequence data; the BACs that are analyzed are restricted to those placed on the FPC map . When several consecutive restriction fragments are present , BACop places a set of contigs , belonging to a BAC , onto the optical contig using boundaries consistent with the upper size range ( 250 kb ) of such clones . This alignment also considers the fragment sizing error model used for alignment of optical and sequence in silico maps [78] . The final placement of optical map contigs onto the maize iMap relies on global considerations of BAC locations on the optical vs . FPC maps ( Figure 2 ) . For example , when both maps have placed BACs showing similar ordering and spacing ( with 20% error allowed ) , alignments are said to be “co-linear . ” Overlaps or gaps are represented on the FPC framework when discordant optical and FPC distances range from ∼200 kb to 2 Mb . When an optical contig aligns to multiple locations on the iMap , the alignment having the greatest number of BACs is selected . BACop placed 91% of the optical contigs ( 60/66 ) onto the 2006 FPC map [37] , with 3 additional optical contigs placed onto FPC contigs that lack chromosome assignments ( Table 1; Figure 3 ) . The total breadth ( 2 , 032 . 42 Mb ) of 60 optical contigs placed on this FPC map ( 1 , 981 Mb ) is slightly larger than its total size . This extra mass accrues from optical contigs that bridge across FPC gaps , and pairs of optical contigs that partially span gaps . At these locations FPC gaps ( reported , or optically revealed ) are apparent because one of the overlapping optical contigs in such pairs has very few if any placed BACs , indicating the presence of a large gap between adjoining FPC contigs , or their incorrect placement . For example , optical contigs OMcontigs_28 and 50 were originally incorrectly placed onto Chr 3 FPC contigs ctg120 and ctg121 . Although these optical contigs overlapped , only OMcontig_28 showed a dense pattern of BACs that aligned to FPC contig ctg121 , but none to the adjoining FPC ctg120 within the overlap region . After realigning each half of OMcontig_28 , the half that hadn't aligned was found to align to the end of the chromosome 5 FPC contig ctg255 ( see Figure 3 ) . This result suggests that either FPC contig ctg121 or ctg255 was incorrectly placed . In the same way , each half of the map contigs OMcontigs_1 , 9 , 12 , 16 , and 17 was also realigned , and this led to improved placements on the iMap . In all , these findings suggest that the current assigned locations of some FPC contigs ( ctg166 , ctg172 , ctg180 , ctg183 , ctg197 , ctg331 , ctg332 , ctg377 and ctg378 ) should be reevaluated . We assessed the accuracy of BACop by analyzing the expected placement of 15 telomeric optical contigs onto the ends of chromosomes on the iMap . Figure 3 indeed shows their placement at chromosome ends: OMcontigs_23 , 28 , 31 , 35 , 38 , 39 , 47 , and 61 are respectively anchored on the rightmost ends of Chrs 9 , 5 , 7 , 8 , 4 , 2 , 3 , and 1 . OMcontigs_7 , 13 , 16 , 20 , and 21 are respectively anchored on the leftmost ends of Chrs 5 , 3 , 8 , 7 , and 10 . Also , OMcontig_51 is anchored on FPC contigs ctg368–370 without covering the leftmost end of FPC contig ctg367 on Chr 9 , and OMcontig_8 is anchored on FPC contigs ctg405–420 without covering the rightmost end of FPC contig ctg421 on Chr 10 . These results suggest that FPC contigs ctg367 and ctg421 should be placed elsewhere , since OMcontigs_51 and 8 have contig “edges” that may represent telomeric regions . The telomeric portion of OMcontig_28 is anchored on FPC contig ctg255 at the rightmost end of Chr 5 , and the other portion of OMcontig_28 is anchored on FPC contig ctg121 , which is placed on the iMap Chr 3 pericentromeric region . Our findings here indicate that FPC contig ctg121 probably should be joined with ctg255 on Chr 5 . We evaluated the quality of available and ongoing maize sequence assemblies by comparing optical contigs completely spanning large “supercontigs” ( pseudomolecules ) from Chrs 1 , 3 and 9 [54] ( Figure 4 and data not shown ) . Our alignments show 9 map segments in common , spanning 2 . 29 Mb ( 29 . 37% ) , between OMcontig_15 and the Zm1S_supercontig ( Chr 1 ) in silico restriction map , and 9 in common between OMcontig_23 and Zm9L_supercontig ( Chr 9 ) covering 3 . 62 Mb ( 54 . 85% ) . However , the Chr 3 finished supercontigs , corresponding to GenBank EF517601 and EF517600 [17] , respectively , showed perfect alignment within OMcontigs_13 and 46 , demonstrating the efficacy of our approach ( data not shown ) . The lack of comprehensive alignment between the optical and the in silico maps for Chrs 1 and 9 pseudomolecules is not surprising because most of the sequenced BACs are in phase 1 assembly , awaiting the ordering and orienting of their associated sequence contigs . Accordingly , gaps of unknown size remain both within and between these nascent sequence assemblies . Based on these alignments of optical contigs vs . pseudomolecules , we characterized many of these gaps and identified issues with orientation . The assembly of the Zm9L_supercontig appears to be superior to that of the Zm1S_supercontig . This view is further buttressed by the higher proportion of phase 2 BAC sequences ( 28/56 ) in the Zm9L_supercontig than in the Zm1S_supercontig ( 14/60 ) . The process of constructing a large sequence pseudomolecule is an iterative one , drawing support for provisional assembly from many sources that guide the serial generation of hypothetical builds and their subsequent validation . As such , we performed a series of optical vs . sequence contig alignments that tracked , guided , and validated the ongoing sequence finishing efforts of a ∼22 Mb sequence pseudomolecule ( FPC ctg182 ) by the Arizona Genomics Institute ( AGI ) . Figure 5 shows two versions of ctg182 , V3 and V7 , aligned to the optical contig—OMcontig_1 . The earliest sequence contig build ( V3 ) contained 12 segments that aligned ( ∼74 . 9% ) with the optical contig , totaling 16 . 30 Mb . Rounds of directed sequence finishing effort led to the construction of the updated build , V7 , which addressed discordances . V7 and the optical contig show an increased alignment of 89 . 6% with 8 larger segments aligning , totaling 19 . 51 Mb . The assembly of the accessioned golden path ( AGP ) involved the merging of 435 correctly ordered FPC pseudomolecules ( 90 . 2 kb–34 . 783 Mb; 2 , 061 Mb total ) to span across the entire maize genome , and it is now known as the B73 RefGen_v1 [53] , [79] . These FPC contig pseudomolecules are constructed from recent sequencing data ( 16 , 848 tiled BACs ) . We facilitated the construction of the AGP by ordering these 435 FPC pseudomolecules , using their alignment to our optical contigs to place them . We uniquely placed 338 of the 435 FPC pseudomolecules onto optical maps; 16 were placed on two optical map contigs , bridging two optical map contigs . Alignments also revealed two possible FPC chimeras ( ctg84 and ctg299; Table S1 ) . The remaining 82 FPC pseudomolecules ( ∼63 Mb ) were not placed on optical contigs due to regions bearing few SwaI sites , or to problems in sequence assembly . Among the 338 uniquely placed FPC pseudomolecules , 65 ( ∼19% ) are either newly placed ( 33; Table S1; blue rows ) or reassigned to amended locations ( 32; Table S1; yellow rows ) . Overall , these results demonstrate the utility of a scalable optical map framework for guiding sequence assembly within a complex genomic environment . The accessioned golden path ( AGP ) ( B73 RefGen_v1 ) from the Arizona Genomics Institute recently released by the maize genome sequencing consortium comprises 10 chromosome-wide “reference chromosomes” ( pseudomolecules ) , and its assembly was guided by several physical maps , including the optical mapping findings presented here ( http://www2 . genome . arizona . edu/genomes/maize_contig_quality_table ) . The B73 RefGen_v1 reference chromosomes represent a unified genomic resource showing chromosome-wide placement of sequence and associated gaps . We provided an independent , optical reference map for this important resource via alignments that comprehensively revealed and sized sequence gaps in FPC pseudomolecules and the B73 RefGen_v1 , which compose reference chromosomes . Local alignment reveals AGP assembly errors characterized as novel gaps , extra and/or missing cuts , and fragment sizing errors . In total , 1 , 102 optical contig segments ( strings of contiguous restriction fragments ) aligned to the B73 RefGen_v1 reference chromosomes ( 1 , 014 . 49 Mb , or ∼50% of AGP [2 , 046 . 35 Mb]; Table 2 ) . The number of optical contig segments that align per chromosome ranges from 74 ( Chr 10 ) to 159 ( Chr 1 ) , and the average map segment size is 937 . 67 kb ( Table 2 ) . The total aligned mass per chromosome varies from 64 . 65 Mb ( Chr 8 ) to 166 . 15 Mb ( Chr 1 ) . The coverage by the aligned map segments for all the maize chromosomes ranges from 37 . 04% ( Chr 8 ) to 59 . 15% ( Chr 4 ) and averages 49 . 58% for all chromosomes ( Table 2 ) . Since the construction of the B73 RefGen is still ongoing , we expected that the optical map: B73 RefGen_v1 alignments would reveal a high level of discordance and an attenuated rate of optical contig alignment . A total of 4 , 465 discordances are identified ( Table S2 ) . These findings include 564 loci with extra sequence data , 829 revealing novel gaps or missing sequences , 2 , 348 misassemblies , 478 additional SwaI restriction sites , and 246 missing SwaI restriction sites . Gaps between adjacent FPC contigs were sized by alignments of optical contigs that span across them . Accordingly , gap size is determined by comparing optical map and B73 RefGen_v1 coordinates across a gap formed between neighboring FPC contigs in the B73 RefGen_v1 reference chromosomes . The FPC contigs do not continuously and seamlessly align to optical contigs since they are constructed from unfinished BACs ( Figure 3 shows optical alignments to FPC contigs ) . Thus we estimated B73 RefGen_v1 gap sizes by considering the pair of coordinates on an aligned optical contig that most closely flank the spanned FPC gap ( Figure S1 ) . More precisely: Gap ( kb ) = [| ( right optical coordinate ) − ( left optical coordinate ) |−| ( right sequence coordinate ) − ( left sequence coordinate ) |]/1000 . In this way , we characterized 263 gaps ( Table S3 ) comprising 44 “negative gaps” ( false B73 RefGen_v1 gaps , or novel sequence ) and 219 “positive gaps” ( confirmed FPC gaps , or unaccounted sequence ) . These 263 gaps were called taking the optical mapping sizing error per restriction fragment into consideration , which is typically +/−5% [80] . However , optical sizing errors can accrue in a complex way across long genomic regions that are spanned by summing consecutive restriction fragments [81]–[82] . As such , 169 of the 263 gap calls were conservatively made when the AGP and optical alignments differences were ≥10% , and the remainder was called below this threshold . Here differences <10% indicate the presence of gaps that were called with less confidence , but their tabulation provides considered targets for sequence bridging and filling . In all , 155 gaps were bridged by optical contigs ( covering 36 . 59 Mb of gaps ) , and an extra 2 . 09 Mb of AGP pseudomolecule sequence was identified .
An optical map was created that spans across ∼91% of the maize ( Zea mays L . ) B73 inbred line ( PI 550473 ) genome , which is a parent of the IBM mapping population . 66 optical contigs are included in this map representing 2 , 103 . 93 Mb of the maize genome decorated by 91 , 453 ordered SwaI restriction sites with accurate physical distances between these sites . On average , there is a SwaI site every 23 kb across the genome , and this restriction recognition sequence “marker” density is far greater than those on genetic ( ∼6 , 000 markers ) and FPC ( ∼24 , 000 markers ) maps [37] ( http://maize-mapping . plantgenomics . iastate . edu/ ) . Because the optical data format is a high-resolution ordered restriction map ( SwaI ) , we were able to anchor and orient FPC-sequence contigs ( http://www2 . genome . arizona . edu/genomes/maize_contig_quality_table ) onto this scaffold . While the immediate utility of the maize optical reference map is as an independent reference for sequence finishing and closing gaps , it will also drive comparative studies for unraveling complex patterns of structural variation as additional inbred lines and cultivars are mapped . Here the optical reference map would serve as a scaffold for future map assemblies enabling rapid discernment of genomic architecture . In this regard , optical mapping may be unique since large ∼500 kb molecules are directly mapped , and this advantage supports scalable genome analysis spanning from a restriction site to multi-megabase-sized regions . The de novo approach that we used to construct the maize optical reference map ensures that it a unique , purely independent resource for sequence assembly and validation . ( The ∼2 . 1 Gb map constructed de novo represents the largest created using single , genomic DNA molecules . ) This physical map is free from common cloning and PCR artifacts , since individual genomic molecules are directly analyzed . These advantages are demonstrated by our comprehensive analysis of several pseudomolecules ( Figure 4 and Figure 5 ) , both published and ongoing , as well as B73 RefGen_v1 reference chromosomes ( Table S3 ) spanning the entire maize genome . Our development of a new algorithm , named BACop ( Materials and Methods; Figure 2 , Figure 3 , and Figure 6 ) , greatly facilitated our ability to analyze and contribute to ongoing sequencing efforts . BACop specifically addressed issues of aligning nascent sequence builds of BAC clones , harboring multiple unordered and unoriented contigs ( averaging 11 per BAC ) , against the optical reference map . Here , BACop was able to link optical contig maps to many unfinished BAC sequences already placed on the FPC map [37] , and to presciently orient and order optical findings across all 10 maize chromosomes . Furthermore , BACop enabled the placement of 60 of the 66 optical contigs onto iMap and identified 12 FPC contigs whose current placement on the iMap requires reevaluation . The calls on 11 out of the 12 FPC contigs identified by BACop for replacements on the iMap ( ctg121 , ctg166 , ctg172 , ctg180 , ctg183 , ctg197 , ctg332 , ctg333 , ctg367 , ctg377 and ctg378 ) were also supported by comparative analysis of optical maps and in silico maps of the FPC contig sequence pseudomolecules ( Table S1 ) . The additional FPC contig identified for replacement by BACop ( ctg421 ) was supported by other sequence markers indicating a new placement abutting ctg90 on chromosome 2 [79] . This BACop algorithm , in combination with optical map data , can also order and orient nascent sequence assemblies . As shown in Figure 6 , many of the unordered and unoriented BAC subcontigs for several clones are nicely placed onto an optical map . Accordingly , BACop provides a useful tool for guiding the ongoing finishing of individual BACs . Given the abundance of maize repetitive sequence , restriction maps directly constructed from ∼500 kb genomic molecules offer many advantages for spanning and structurally characterizing heterochromatic regions . This advantage is evidenced by the long contigs ( 3 . 64 Mb to 100 . 76 Mb ) within the optical reference map , averaging 31 . 88 Mb in length . In comparison , the May 2006 maize FPC map comprises 721 FPC contigs averaging ∼2 . 98 Mb in length with a total mass of 2 , 150 Mb ( 300/721 , or 163 . 7 Mb , are unassigned ) [37] . Long optical contigs offer unique benefits especially when they span across genomic regions sparsely populated by markers , enabling structural insights to be drawn in these regions . In part , these advantages have characterized gaps and sequence misassemblies and reordered 19% of the FPC pseudomolecules ( Table S1 ) , which was persisted into the current B73 RefGen_v1 sequence for the maize genome . Maize centromeres have been mapped to regions with flanking molecular markers using many different approaches [83]–[88] . However , for some maize chromosomes such as Chrs 1 , 3 , and 6 , the proposed centromeric locations differ among different mapping methods; while for other chromosomes - Chrs 2 , 4 , 9 , and 10 - there is a consensus running across different mapping techniques [84] . Recently , maize centromeres were located on the B73 RefGen_v1 sequence using centromeric markers [53] derived from numerous sources: transposon display , repeat junction , centromere repeat , and chromatin immunoprecipitation data [53] . Accordingly , we located centromeric loci assigned to Chrs 1 , 2 , 5 , 6 , 8 , 9 , and 10 around gaps in the optical contigs: [CEN1 ( OMcontig_3 and 69 ) ; CEN2 ( OMcontig_2 and 41 ) ; CEN5 ( OMcontig_22 and 42 ) ; CEN6 ( OMcontig_65 and 36 ) ; CEN8 ( OMcontig_16 and 4 ) ; CEN 9 ( OMcontig_55 and 17 ) ; CEN10 ( OMcontig_49 and 17 ) ] ( data not shown ) . Unfortunately , the flanking contigs did not fully span any centromeric regions; however , these optical contigs did structurally characterize several pericentric regions . Although we have demonstrated here that optical mapping offers numerous benefits for physical mapping and genome sequence assembly , its utility would be appreciably extended when combined with next generation sequencing . Genome analysis approaches are now rapidly evolving and tracking the increasingly cost-effective capabilities offered by next generation sequencing . Next generation sequencing approaches using single molecule libraries are now tackling the analysis of complex genomes [89]–[90] , but they do not offer data sets competent for de novo assembly because of modest read lengths and errors . As such , new sequencing strategies must be developed for wheat and other complex crop genomes that effectively seize the new opportunities enabled by next generation sequencing . To this end , we propose the proactive use of optical mapping data for sequence assembly [91] . The combination of long-range ( optical ) and nucleotide-level ( next generation ) data sets , both generated directly from genomic molecules , may prove to be a cost-effective approach - especially when new algorithms are developed that intimately comingle both data sets during the assembly process .
Maize kernels ( inbred line B73 , PI550473 ) , obtained from the USDA-Agriculture Research Service North Central Regional Plant Introduction Station ( Iowa State University , Ames , IA 50011-1170 ) , were washed in 10% Clorox bleach for 10 min , rinsed in sterile water ( 3× , ∼3 min per wash ) , and germinated on moistened brown paper towels in a dark , moist chamber at 30°C for 12 days . Residual ungerminated seeds were removed from maize sprouts prior to nuclei isolation . The procedures for isolation of nuclei and storage have been described previously [75] . Prior to use , isolated nuclei were washed 2× with fresh Dulbecco's PBS ( 1 . 54 mM KH2PO4 , 155 . 17 mM NaCl , 2 . 71 mM Na2HPO4 , pH 7 . 2 ) to remove glycerol . Rapid DNA concentration assays were conducted by lysing small aliquots of nuclei in TE ( 10 mM Tris-Cl , pH 8 . 0 , 1 mM EDTA ) with 1 mg/ml proteinase K , and adenovirus DNA added at 25 pg/µl ( internal sizing standard; Invitrogen , Carlsbad , CA ) , followed by mounting , restriction digestion , staining and imaging as previously described . Appropriate dilutions for mapping ( optimized to minimize molecular crossovers ) were made by adjusting the amount of isolated nuclei in the lysing solution ( TE with 1 mg/ml proteinase K , 25 pg/µl adenovirus DNA in TE ) , by slowly pipetting up and down several times using a wide-bore pipette tip; samples were incubated at 65°C for 1 hr and at 37°C overnight . Samples were mounted onto optical mapping surfaces and imaged by fluorescence microscopy to assess DNA integrity and concentration of both genomic and reference standard DNA molecules . Surface preparation was done as previously described [71]–[72] . Briefly , glass cover slips ( 22×22 mm , Fisher's Finest ) were cleaned by boiling in Nano-Strip ( Cyantek Corp . , Freemont , CA ) , acidified by boiling in concentrated HCl , extensively rinsed with high purity water and ethanol under sonication , and derivatized using trimethyl and vinyl silanes to confer a positive charge and the means to crosslink the acrylamide overlay to the surface . Surfaces were evaluated by mounting lambda DASH II bacteriophage DNA ( Stratagene , La Jolla , CA ) and digesting them with 40 units of SwaI , diluted in 100 µL of digestion buffer containing 0 . 02% Triton X-100 at room temperature to determine the optimal digestion time ( 30 min to 2 . 5 hrs ) . Genomic DNA molecules ( ∼400–500 kb ) premixed with the adenovirus DNA sizing standards were deposited as stripes on derivatized glass surfaces using a silastic microchannel system [92] . A fully automated image acquisition and processing system collected data and compiled large files consisting of an ordered restriction map for each genomic DNA molecule . All microscope and camera functionalities and machine vision processes are fully automated and controlled by computer software . Detailed procedures were previously described [74]–[76] , [92] . With a raw map data set of >2 million maps , a divide and conquer strategy for optical map assembly was needed to deal with the severe computational load through parallel processing . We previously used this approach for the assembly of genome maps spanning the rice as well as the Leishmania major genomes [74]–[75] . Briefly , the map data set was divided into smaller sub-data sets ( ∼30 , 000 single molecule optical maps ) allowing efficient parallel assembly [93]–[96] , over 2–3 days , without taxing computer memory limits . The consensus maps from all contig assemblies were reassembled together for identifying redundant contigs and for merging overlapping optical consensus maps . After this reassembly process , a unique set of optical consensus maps was identified as “seed” maps for initiating iterative assembly . Iterative assembly consists of cycles of pairwise alignment [78] of the entire map data set against seed maps , followed by the contiging of these aligned single molecular optical maps for extending and refining seed maps in each subsequent cycle . Cycles of iterative assembly broaden and increase the coverage depth of nascent contigs . Consensus maps are then stripped from the newly formed contigs as updated seed maps after further processing . The pairwise alignment phase extracted multiple high-scoring alignments based on the efficient linear scaling approach of Huang and Miller [97] , and the confidence scores ( p-values ) were generated using an approach similar to that used by Waterman and Vingron [98] . Updated consensus maps were assembled to identify redundancy and to merge overlapping consensus maps . This process identified seed maps for the next iteration , and this iteration process was repeated typically more than ten times until the optical map contigs no longer grew . Large contigs ( >10 Mb in breadth ) also present computational challenges . For these contigs , iterative assembly considers and augments only the terminal 40 restriction fragments . About 85% of the maize genome comprises extensive families of repetitive sequences . Consequently , multiple contigs emerge from the sequence assembly of a single BAC , which are also unordered and unoriented . In order to integrate our optical map with the iMap , we developed a new algorithm —“BACop” — that utilizes unfinished BAC sequence data . The algorithm for anchoring the optical maps on the maize genetic-physical ( FPC ) map precedes in four distinct steps: i ) matching restriction fragments between the optical map and the in silico restriction fragments from the sequence contigs of the BACs , ii ) determining locations of all the BAC sequence contigs along the optical map , iii ) anchoring the BACs on the optical maps , and iv ) filtering and combining the alignments of BACs in the FPC map and optical map to find the most significant ones ( Figure 2 ) . The first step compares individual restriction fragment sizes from the optical map contigs with the in silico restriction fragments of the sequence contigs of the BACs . A fragment from the optical map assembly of size X and an in silico restriction fragment of size Y match if |X−Y|/σ√Y< = k , for parameters σ and k based on the statistical model developed by Valouev et al . 2006 [78] . Once the matching fragments have been determined , the BACs are located on the optical map assembly by examining each BAC's in silico restriction fragments . Approximate locations are determined by a filtration method that selects candidate regions on the optical map assembly that a BAC can align to , based on the matching fragment density . The approximate locations are further screened to produce a feasible alignment of a BAC's restriction fragments to the restriction fragments on the optical map assembly . Since each BAC is shotgun sequenced , multiple sequence contigs can result since the orientation and order of sequences are unknown . A feasible alignment must preserve the order of the in silico restriction fragments from within the same sequence contig but is allowed to match fragments from different sequence contigs in any order and orientation . A match graph is constructed from a candidate region on the optical map assembly that the BAC can align to using the matching restriction fragments between the two as nodes within the graph . A traversal through the match graph induces an alignment of the BAC to the optical map assembly from the nodes representing matching fragments along the traversed path . The graph traversal resembles a branch and bound algorithm that exhaustively enumerates all feasible alignments and selects the best one . After all possible locations of the BACs are determined , the optical contig is then aligned to the FPC map . The FPC map gives the relative positions of the BACs within each FPC contig as well as the positions and order of the FPC contigs with respect to each other . Dynamic programming is used to align the optical map contigs to the FPC contigs by scoring for matching BACs along the optical contig that respect the order and location of BACs along the FPC map . A scoring scheme that weighs for higher quality BACs based on sequencing status and for BACs with greater fragment density is used . A fudge factor is applied when examining the locations of the BACs specified by the FPC map since they are approximate . Gaps between FPC contigs are specified using lower and upper bounds . The alignment is evaluated based on the number of BACs that are scored within the alignment region to remove spurious alignments according to a set threshold . The threshold is adjusted to allow for alignments in regions where there is sparse sequence data resulting in a lower number of usable BACs to align to . All of the alignments made with different thresholds are then collected , and the best alignments are selected according to coverage of the optical map assembly and number of aligned BACs . | The maize genome contains abundant repeats interspersed by low-copy , gene-coding sequences that make it a challenge to sequence; consequently , current BAC sequence assemblies average 11 contigs per clone . The iMap deals with such complexity by the judicious integration of IBM genetic and B73 physical maps , but the B73 genome structure could differ from the IBM population because of genetic recombination and subsequent rearrangements . Accordingly , we report a genome-wide , high-resolution optical map of maize B73 genome that was constructed from the direct analysis of genomic DNA molecules without using genetic markers . The integration of optical and iMap resources with comparisons to FPC maps enabled a uniquely comprehensive and scalable assessment of a given BAC's sequence assembly , its placement within a FPC contig , and the location of this FPC contig within a chromosome-wide pseudomolecule . As such , the overall utility of the maize optical map for the validation of sequence assemblies has been significant and demonstrates the inherent advantages of single molecule platforms . Construction of the maize optical map represents the first physical map of a eukaryotic genome larger than 400 Mb that was created de novo from individual genomic DNA molecules . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/genome",
"projects",
"genetics",
"and",
"genomics/genomics",
"genetics",
"and",
"genomics/plant",
"genomes",
"and",
"evolution",
"genetics",
"and",
"genomics/chromosome",
"biology"
] | 2009 | A Single Molecule Scaffold for the Maize Genome |
Understanding the medical effect of an ever-growing number of human variants detected is a long term challenge in genetic counseling . Functional assays , based on in vitro or in vivo evaluations of the variant effects , provide essential information , but they require robust statistical validation , as well as adapted outputs , to be implemented in the clinical decision-making process . Here , we assessed 25 pathogenic and 15 neutral missense variants of the BRCA1 breast/ovarian cancer susceptibility gene in four BRCA1 functional assays . Next , we developed a novel approach that refines the variant ranking in these functional assays . Lastly , we developed a computational system that provides a probabilistic classification of variants , adapted to clinical interpretation . Using this system , the best functional assay exhibits a variant classification accuracy estimated at 93% . Additional theoretical simulations highlight the benefit of this ready-to-use system in the classification of variants after functional assessment , which should facilitate the consideration of functional evidences in the decision-making process after genetic testing . Finally , we demonstrate the versatility of the system with the classification of siRNAs tested for human cell growth inhibition in high throughput screening .
Genetic tests , that aim to identify disease-associated germline variants in the genome of patients and relatives , have greatly expanded these last years , together with the number of predisposing genes scrutinized [1] . Genetic tests are proposed by genetic counselors to identify the carriers of genetic variants and to define appropriate clinical follow-ups and treatments for these carriers . The detection of a variant can have severe psychological and physical consequences for the tested patients , depending on whether the variant is known to be pathogenic ( associated with disease development ) , neutral ( not related to disease development ) or of unknown significance ( VUS ) . Thus , clinical decision-making after genetic testing requires the establishment of reliable variant classifications . The best support is to use methods that attribute a probability of pathogenicity for each variant identified . Because genetic/epidemiological methods , such as co-segregation , case-control , co-occurrence and familial data analyses , provide such probabilities [2] , they remain the gold standard in clinical decision-making after genetic testing ( see an example in S1 Table ) . However , genetic/epidemiological methods are time consuming , as they require a substantial amount of observations . Moreover , they are unsuitable for a large number of variants identified , for instance when the number of known carriers is rare . As genetic tests are evolving towards the use of multi-gene panels , whole exome and whole genome sequencing [3] , the number of VUS detected is inevitably increasing [1] , which stresses the need to improve variant classification [3] . Functional assays have been designed to circumvent the limitations of genetic/epidemiological methods . The generic "functional assay" term refers to in vitro and in vivo systems , able to classify VUS by assessing their influence on protein function or conformation [4] . Functional assays have been widely developed for genes involved in cancers [5] and BRCA1 has become the leading gene analyzed , with 23 different assays proposed , to date [6] . However , despite the genuine interest for strategies that alleviate the limitations of genetic/epidemiological methods , the main challenge of functional assessment remains in its inclusion into clinical decision-making . Indeed , most of the functional assays lack statistical validation [4] . Moreover , analyses are usually based on visually defined cut-offs [6] . Finally , except in rare cases [7 , 8] , the resulting variant classifications lack the probability of pathogenicity provided by genetic/epidemiological methods . Here , we used experimental as well as computational approaches to overcome these limitations . We evaluated the clinical utility of four different BRCA1 functional assays , designed in yeast cells , by assessing 40 BRCA1 missense mutations , previously classified by genetic/epidemiological methods . To interpret these results , we developed a novel approach , referred to as "Mann-Whitney-Wilcoxon ( MWW ) method" , that defines a non-arbitrary best cut-off value between the neutral and pathogenic variants and that refines variant ranking in data from functional assays . We also developed a computational system that transforms the dual classification between "pathogenic" or "neutral" , provided by the non-arbitrary best cut-off , to a probabilistic classification adapted to clinical decision-making . This system of classification , referred to as "probability system" , uses the fluctuation of the best cut-off to derive probabilities of pathogenicity for each assessed variant . We show the benefit of our computational model , coupling the MWW method and the probability system , using the experimental data from the four BRCA1 functional assays and using theoretical simulations . We also illustrate that our model is adapted to experimental systems far beyond the genetic variant assessment , with the probabilistic classification of small interfering RNAs ( siRNAs ) tested for human cell growth inhibition in high throughput screening .
The colony size assay is a BRCA1 functional assay , that has been designed in the yeast model organism , which allows rapid , large-scale and cost-effective variant assessment [9] , but has never been subjected to clinical validation yet . In this functional assay , expression of the full length wild type ( WT ) BRCA1 protein in yeast , induces a growth defect [9–11] . Indeed , after 63 hours of growth on an agarose plate , a single yeast cell gives rise to a colony varying between 5 , 000 and 21 , 000 cells ( Fig 1A , BRCA1 ) , while colonies reach several millions of cells without protein expression ( Fig 1A , Vector control ) . To ascertain the utility of this assay in clinical medicine , we selected 40 BRCA1 missense mutations , according to their neutral or pathogenic classification by genetic/epidemiological methods ( S1 Fig and S2 Table ) . We confirm that pathogenic missense mutations restore the proliferation rate of yeast cells [10 , 11] . Indeed , pathogenic mutations have a global tendency to give rise to the biggest colonies , while colony sizes arising from neutral mutations remain close to those of the WT BRCA1 reference ( Fig 1A ) . However , the Colony Size assay does not fully discriminate between pathogenic and neutral mutants . Indeed , variant medians appeared to continuously decrease from M1689R ( highest median ) to V1804D ( lowest median ) , without clear gap between the pathogenic and neutral regions . Moreover , the neutral M1652T mutation is clearly within the pathogenic sector and the pathogenic R1699W mutation slightly overlaps the neutral region . In such situations , it is critical to have a sound evaluation of the sensitivity , specificity and accuracy of the assay ( see the definitions in the S1 Text ) , which depends on a non-arbitrary and optimal cut-off setting . The standard method is based on the Youden's index , a classical approach to compute the sensitivity and specificity in a dataset . Using this , the cut-off of 17 , 910 cells per colony gives the best combined sensitivity and specificity , with 96% ( 24/25 ) and 93% ( 14/15 ) respectively ( Table 1 and S2A Fig ) . In total , 95% ( 38/40 ) of the mutations are correctly classified . The M1652T neutral mutation is misclassified as pathogenic and the pathogenic R1699W mutation is misclassified as neutral ( S3 Table ) . From now on , we refer to "experimental best cut-off" , "experimental sensitivity" , "experimental specificity" and "experimental accuracy" as the best cut-off , sensitivity , specificity and accuracy obtained from the experimental data . The disadvantage of the standard method is that mutations are characterized by a single value , here by the median of colony sizes , which can lead to paradoxes in the mutant classification . For instance , the neutral I1858L mutation displays a median of cells per colony higher than the median of the neutral T1720A mutation . Thus , in the mutant ranking , I1858L is closer to the pathogenic group of mutations than T1720A ( arrows in Fig 1A ) . However , T1720A has three values out of nine over the experimental best cut-off , which are thus in the pathogenic area , while I1858L has none ( S3A Fig ) . Therefore , in terms of dispersion range , T1720A could be considered as "more pathogenic" than I1858L . To overcome such paradoxes in variant classification , we developed a nonparametric approach to define the best non-arbitrary cut-off value , that takes into account more information from distributions than the median value alone . This method is based on the MWW test [12–14] . Since the p value of this test provides a quantification of the overlap between two distributions ( S4 Fig ) , we compared each mutant distribution to the WT BRCA1 distribution . The p values obtained defined relative positions of the mutant distributions using the WT BRCA1 distribution as a reference position ( Fig 1B and S4 Table ) . Contrary to the standard method described above , the cut-off used to compute the sensitivity and specificity parameters is a p value . Any mutant with a p value below the p value cut-off , indicates a mutant classified as pathogenic . In contrast , a mutant distribution with a p value over the p value cut-off is considered as neutral . Strikingly , the MWW method solves the paradoxes observed with the standard method , since T1720A is closer to the pathogenic group of mutations than I1858L ( arrows in Fig 1B ) . Moreover , the experimental sensitivity and specificity remains unchanged ( Table 1 and S2E Fig ) . This confirms that the M1652T and R1699W mutations cannot be correctly classified by the Colony Size assay , even when using more information from the experimental data than the variant medians alone . However , it also emphasizes that the variant classification , provided by the MWW method , does not diminish the high sensitivity and specificity of the assay . From this , we conclude that the MWW method is a reliable alternative to the standard method to define a non-arbitrary cut-off in data from functional assessments . Recently , two-component models have been proposed for the probabilistic classification of variants based on functional assessment . These parametric models require the normal distribution of the neutral and pathogenic values [7 , 8] . However , as shown in S5A Fig , the Colony Size assay is poorly compatible with such models , due to the bimodal distribution of the pathogenic values . Therefore , we designed an alternative nonparametric and more versatile system of classification . This system is based on the fact that the best cut-off is a random variable that fluctuates , depending on the experimental values . We asked what the variant classification would be , using the Colony Size assay , taking this fluctuation into account . For this , we performed sampling with replacement ( bootstrap ) of the colony size values , by randomly choosing 9 values among the 9 from each mutant , and 36 values among the 36 from the BRCA1 reference control . Next , using this new set of sampled data , we applied the standard or MWW method to obtain the best cut-off . We repeated this procedure a large number of times , which allowed us to define a best cut-off distribution for the standard and MWW methods ( S5 Table ) . We also used a third method , referred to as "standard with reference method" . It is similar to the standard method , except that the best cut-off distribution obtained includes the fluctuation of the WT BRCA1 reference , as explained in the S1 Text . Notably , the standard with reference method allows an additional comparison with the MWW method , which also includes the fluctuation of the WT BRCA1 reference . Finally , we designed the probability system of classification . This system allows to assign a probability of pathogenicity to each assessed variant , using the best cut-off fluctuation ( Fig 2A and S6 Fig ) . The rationale is that the farther a variant is from the core of the best cut-off fluctuation , the more robust is its classification as either pathogenic or neutral . A probability close to 1 indicates that the variant can be classified as pathogenic , with a low risk of misclassification as neutral due to the fluctuation of the best cut-off . A probability close to 0 indicates that the variant can be classified as neutral , with a low risk of misclassification as pathogenic due to the fluctuation of the best cut-off . Finally , a probability of 0 . 5 designates no preferential classification as either neutral or pathogenic ( variant completely unknown ) . With such probabilities , the five-class nomenclature proposed by Plon et al [26] . ( S1 Table ) can be directly applied to functional assays . Probabilities obtained for the Colony Size assay are shown in Fig 2B . Strikingly , a level of uncertainty was generated , notably with variants classified as "uncertain" ( class 3 ) . This highlights the critical influence of the best cut-off fluctuation in variant classification . In addition , the MWW method exhibits the best accuracy , with 37/40 mutations correctly classified versus 36/40 for the standard and standard with reference methods . When including the number of misclassified mutations , the MWW method shows a balance of 35 mutations , ex-aequo with the two other methods ( S6 Table ) . Altogether , these results confirm the possibility to use the MWW method in variant classification . In addition , the probability system seems to be an effective and simple way to obtain a probabilistic classification of variants in functional assessment . We validated three other functional assays , by assessing the same 40 mutations used in the Colony Size assay . The Liquid Medium assay monitors the growth defect of yeast cells expressing BRCA1 ( S7 and S8 Figs ) , as in the Colony Size assay , but in liquid instead of solid medium [11] . The Spot Formation assay is derived from the observation that the BRCA1-mCherry fusion protein accumulates in a single aggregate in the nucleus of yeast cells . This aggregate is referred to as "spot" due to its visual signature using fluorescent microscopy . We previously showed that pathogenic missense mutations decrease the proportion of cells showing one spot [11] . Here , we confirmed this effect ( S9 and S10 Figs ) . The last assay tested was the Yeast Localization assay . Whereas cytoplasmic spots are rare in yeast cells expressing the WT BRCA1 protein , this event has a tendency to increase in the presence of pathogenic mutations [11] . Here , we confirmed this effect ( S11 and S12 Figs ) . However , albeit promising , none of these three assays provided a better discrimination than the Colony Size assay , to distinguish between pathogenic and neutral variants . This was notably shown by the experimental sensitivity and specificity computed ( Table 1 and S3 Table ) . We took advantage of the experimental differences among the four assays ( recapitulated in S7 Table ) to detect potential flaws in the MWW method . In contrast , the MWW method constantly overcomes the incoherent ranking generated by the standard method ( see examples in S7 , S9 and S11 Figs ) . This is achieved without reducing the experimental accuracy compared to the standard method ( Table 1 ) , except for a minor decrease in the Liquid Medium assay ( 88% versus 90% ) . Also , no flaws were detected in the probability system , which would result in an unexpected high level of misclassifications ( Fig 2B ) . Interestingly , accuracy of the MWW method is globally better than in the standard or standard with reference method , with the best accuracy of 93% in the Colony Size assay , 83% in the Spot formation assay , and with the best ex-aequo accuracy of 73% in the Yeast Localization assay ( Fig 2B ) . Variant misclassification was slightly higher in the MWW method , compared to the two other methods , with one more misclassification in the Colony Size and in the Spot Formation assays , one less in the Liquid Medium assay , and ex-aequo in the Yeast Localization assay ( "Total number of variants misclassified" column in S6 Table ) , even if the balance between accuracy and misclassification maintains the MWW method as the best one , ex-aequo with the standard method ( "Balance" column in S6 Table ) . Finally , contrary to the MWW method , the standard method suffers from a lack of sensitivity in the Yeast Localization assay , since none of the pathogenic mutations are classified as class 5 ( Fig 2B ) . Overall , the analysis of four functional assays did not reveal any major flaw in the probability system of classification . In addition , the results obtained with the MWW method confirm the possibility to classify variants using more information from the variant distribution than the median value alone . To complete the detection of potential flaws in our classification model , we analyzed theoretical situations . A reference situation was designed , similar to that in the Colony Size assay ( S8 Table ) . Next , different parameters were scrutinized: the position of the pathogenic mutations ( S13 Fig ) , neutral mutations ( S14 Fig ) , or WT BRCA1 reference ( S15 Fig ) , the initial sensitivity and specificity of the assay before using the probability system ( S16 Fig ) , the number of neutral and pathogenic variants used ( S17 Fig ) , the number of values in the variants and in the WT reference distributions ( S18 Fig ) , and the range of the variant and WT reference distributions ( S19 Fig ) . Results are recapitulated in S9 Table and summarized in Table 2 . The standard with reference method shows strong usage limitations , notably when the WT reference exhibits a negative median or a median close to zero ( Table 2 and S15E Fig , middle panel ) . Interestingly , the MWW method is not affected by such situations . The main limitation detected is an extreme situation in which the WT reference distribution falls outside of the range of the neutral and pathogenic distributions ( e . g . , S15A Fig , left panel ) , which impairs the sensitivity of the probability system of classification ( Table 2 and S15E Fig , right panel ) . Except for this extreme situation , we confirm the efficient behavior of our classification model , coupling the MWW method and the probability system: ( 1 ) when the pathogenic and neutral distributions are strictly identical , all the mutations are classified as class 3 ( Table 2 and S13D Fig , right panel ) , ( 2 ) the sensitivity and specificity of the probability system of classification increase when pathogenic mutations move away from the WT BRCA1 reference distribution ( S13D Fig , right panel ) , and ( 3 ) when pathogenic mutations are contaminated by neutral mutations ( experimental specificity reduced ) , the sensitivity of the probability system of classification is decreased ( Table 2 and S16E–S16G Fig , right panel ) , and vice versa . This last result is an important criterion for classification , since unknown mutants that would be located in a pathogenic region containing neutral mutations , could not be formally classified as pathogenic . Therefore , it is noteworthy that the experimental sensitivity and specificity values are taken into account by our classification model . Interestingly , the model is poorly sensitive to the number of neutral or pathogenic mutations used to validate a given assay ( S17E–S17G Fig , right panel ) , as long as the number of values in the dataset is high enough ( S18E–S18G Fig , right panel ) . Supplemental information is provided in the S1 Text . This notably includes an extensive analysis of the best cut-off fluctuation , which explains the lack of sensitivity of the standard method , mentioned above in the Yeast localization assay ( Fig 2B ) and also shown in theoretical situations ( see the legend of S13 Fig ) . It also contains specific procedures for variant classification ( e . g . , Bayesian inference , combination of functional results , assessment of VUS ) , as well as procedures to fit the proposed model to other situations . It finally includes the ProClass toolbox that generates the probabilistic classification of variants , adapted to most kind of functional assays . We wondered if the classification model developed for genetic variants could be easily extended to other decision-making situations . The analysis of theoretical situations showed that variant classification remains accurate when only one neutral and one pathogenic variant are available ( S17E–S17G Fig ) . This indicates that the fluctuation of the best cut-off supports decision-making in situations represented by a limited number of positive and negative controls . To confirm this , we analyzed data from 406 genes targeted by small interfering RNAs ( siRNAs ) , screened for their capability to inhibit the proliferation of a human prostate tumoral cell line ( Fig 3 ) . The "No siRNA" control , the siKIF11 positive control and the siGOLGA2 and siGL2 negative controls were treated as WT reference , pathogenic and neutral variants , respectively . Structure of the data is reported in S7 Table . As in the BRCA1 functional assays , the experimental accuracy was not impaired using the MWW method , compared to the standard method ( Table 1 ) . In addition , no flaws were detected in the probability system , since the accuracy remained at 1 , whatever the standard , standard with reference or MWW method used ( Fig 3C ) . Finally , the advantage of the MWW method is again highlighted in the final classification of the screened siRNAs . Indeed , in the siRNA ranking , based on the median values , siGTSE1 is closer to the negative controls than siITGA2 ( Fig 3A ) . By taking the distribution of these two siRNAs into account , the MWW method switches their ranking position ( Fig 3B ) , so that siGTSE1 is finally classified as "unclear effect on cell growth inhibition" ( Fig 3C , MWW method ) , instead of "no cell growth inhibition" ( Fig 3C , standard and standard with reference methods ) . Thus , this demonstrates that our probabilistic model is also adapted to the classification of experimental data far beyond the functional assessment of genetic variants .
We provide the statistical validation of four BRCA1 functional assays , as well as a classification model that facilitates the incorporation of functional assay results into clinical decision-making . The probabilistic model is based on the fluctuation of the best cut-off , which is driven by the fluctuation of the experimental data . Thus , the variant classification provided reflects the robustness of a cut-off-based decision-making towards data fluctuation . The model has the advantage to be nonparametric , easy to handle and easy to adapt to most kind of functional assays . Moreover , among the variants incorporated in functional assays , the model only depends on those previously classified by genetic/epidemiological methods as pathogenic or neutral . It is not influenced by unknown variants , meaning that the subsequent incorporation of unknown variants in a functional assay does not require a new analysis of the best cut-off fluctuation . These features of our model contrast with parametric models , proposed for variant classification [7 , 8] . We achieved a widespread analysis of the best cut-off fluctuation dedicated to decision-making ( completed in the S1 Text ) . This analysis is focused on the classification of genetic variants , but it is also valid for other decision-making situations compatible with our classification model , such as high throughput siRNA screenings . Using many different kinds of data structures ( four BRCA1 functional assays , one siRNA screen and 93 theoretical situations ) , three different methods of best cut-off fluctuation were scrutinized: the standard , the standard with reference and the MWW methods . From this study , we conclude that the standard with reference method is poorly compatible with a versatile classification model , due to important lacks of accuracy when the WT reference exhibits a negative median or a median close to zero ( S15E Fig , middle panel ) . The standard method has the advantage to support decision-making in experimental situations devoid of a WT reference . The MWW method has the advantage to use more information from the distribution of the classified elements than the median value alone . This refines the ranking and the final probabilistic classification . Contrary to the standard method , the MWW method is adapted to experimental situations in which the neutral and pathogenic variants ( or the negatives and positives controls ) are represented by a single value if the WT reference encompasses a significant number of different values ( S18E Fig , compare the left and right panels for the mutant with one value ) . However , the MWW method is poorly adapted to experimental situations where the WT reference distribution is more or less outside of the range of the neutral and pathogenic distributions ( e . g . , S15A Fig , left panel ) . Thus , we propose to prioritize the MWW method if the data are compatible with this method , notably if the WT reference is well embedded in the neutral/negative distributions , or if the WT reference is between the neutral/negative and the pathogenic/positive distributions , and to use the standard method otherwise . The different methods are proposed in the ProClass toolbox available online ( see the S1 Text ) . Interestingly , none of the four yeast assays is able to correctly classify the R1699W pathogenic and the K45Q neutral variants . Pathogenicity of R1699W has been long-established in independent studies , using different genetic/epidemiological methods [15–17] , confirming that yeast cells are unable to detect the deleterious impact of R1699W [18] . This emphasizes that the mechanism of R1699W , leading to tumor development , is different from the other pathogenic missense variants of BRCA1 . It is probably related to a protein-binding defect without major BRCA1 structural destabilization [19] . The classification of K45Q has been established by a single epidemiological study [20] , with little evidences of neutrality ( e . g . , probability of being pathogenic of 11% by family history prediction ) . However , the absence of any functional impact has been confirmed in three different functional assays using mammalian cells [21 , 22] , which stresses a specific effect of K45Q in yeast cells , that remains to be explained . Finally , this work showed that the yeast organism can be used to classify variants positioned in both Nter and Cter parts of BRCA1 . Among the four assays analyzed , the Colony Size assay is the most accurate ( 93% ) and the most robust to data fluctuation ( one class 3 variant ) . The Liquid Medium and Yeast Localization assays may also be attractive for diagnosis due to the absence of false negative results detected , notably when using the MWW method . Interestingly , the Yeast Localization assay allows the identification of pathogenic variants that delocalize the BRCA1 protein into the cytoplasm . If confirmed in human cells , this assay could define subcategories in the pathogenic variants of BRCA1 , based on different cellular mechanisms leading to tumor development .
All plasmids are derived from pJL48 [11] , a modified version of pESC-URA ( Agilent Technologies ) , in which the MYC epitope has been removed by SalI-XhoI digestion and vector ligation . In this plasmid expression of the cDNA is controlled by the GAL1 promoter , inducible by galactose and repressed by glucose . The backbone of the human BRCA1 ( MIM# 113705 ) cDNA used , corresponds to the AY888184 . 1 GenBank sequence with a TGA stop codon instead of TAG . To facilitate the cloning of BRCA1 missense mutations , silent mutations were inserted in the cDNA to generate 4 new restriction sites: SalI ( c . 1020A>C + c . 1023T>C ) , AvrII ( c . 4662A>T ) , FseI ( c . 4839T>G + c . 4842A>G + c . 4845T>C ) and XhoI ( c . 5502C>T ) . Of note , the WT BRCA1 and BRCA1-mCherry plasmids used in this study ( pPT60 and pPT63 respectively , see S10 Table ) are different from the pJL45 and pGM40 plasmids , used in our previous publication [11] , by the addition of the 4 restriction sites . The 40 missense mutations were generated by targeted mutagenesis ( Genscript Company , Piscataway , NJ , USA ) on intermediate plasmids . Next , we inserted the mutated cDNA fragment into the pPT60 and pPT63 plasmids by a single digestion—ligation step . All resulting plasmid constructs were verified by sequencing the promoter , the full cDNA and the terminator . Transformation of the Saccharomyces cerevisiae haploid BY4741 or YKR082W-GFP strains were performed as previously described [11] . The strains generated are referenced in S11 Table . To facilitate the description , we referred to the cells transformed with pESC-URA as the name of the cDNA inserted into the plasmid . Thus , "BRCA1" refers to yeast cells transformed with the plasmid containing the WT BRCA1 cDNA; "M18T" refers to yeast cells transformed with the plasmid containing the M18T mutated version of the BRCA1 cDNA; and "vector" refers to yeast cells transformed with the same plasmid without inserted cDNA . Three independent transformants per strain , also referred to as "clones" , were selected after each transformation . We observed that lithium acetate transformation can result in diploidisation of haploid cells . To control this , the ploidy of each clone was verified by FACS analysis , using the yeast strain BY4741 ( haploid ) and BY4743 ( diploid ) as a control . Next , in the different assays , cells were grown in glycerol-lactate medium ( GL-URA ) as previously described [11] . Addition of galactose in the medium ( GAL ) induced the expression of BRCA1 , while addition of glucose ( GLU ) strongly repressed the expression of BRCA1 . Cells were grown in log phase in YPD medium ( 1% yeast extract , 2% peptone , 2% dextrose , 60 μM Adenine , 8 μM NaOH ) . 107 cells were collected and put at 4°C to block the cell cycle . Cells were centrifuged at 4°C and resuspended in 70% ethanol . After 1 hour incubation at room temperature ( RT ) , cells were centrifuged and resuspended in freshly made sodium citrate [50 mM] pH7 . Sonication was performed to dissociate cell aggregates ( Vibracell and probe CV33 ( Bioblock Scientific , Illkirch , France ) , pulse 30% , time 15 seconds ) . Cells were centrifuged and resuspended in sodium citrate [50 mM] pH7 + RNAse A [0 . 25 mg/ml] . After 1 hour incubation at 50°C , cells were centrifuged , resuspended in sodium citrate [50 mM] pH7 + Propidium Iodide [16 μg/ml] and analyzed using an Accuri ( BD Bioscience , San Jose , CA , USA ) . This assay was previously named "small colony phenotype" ( SCP ) assay . The method already published [11] was slightly improved as follows: ( 1 ) GL-URA+galactose and GL-URA+glucose plates were incubated 63 hours and 50 hours respectively ( instead of 52 hours ) , and ( 2 ) the biggest colony of each plate , representing the size of at least five other colonies on the plate , was chosen for cell counting . This prevents the choice of rare but extremely big colonies ( outliers ) . For the simultaneous assessment of 10 variants by a single technician , the time required between the delivery of the intermediate plasmids ( see above ) and the final results is 20 days . This method already published [11] was slightly improved as follows: during glucose induction , cells were diluted at 0 . 5×106 cells/ml ( instead of 106 cells/ml ) for the 15 hour culture time at 30°C . Galactose induction conditions remained as before . For the simultaneous assessment of 10 variants by a single technician , the time required between the delivery of the intermediate plasmids ( see above ) and the final results is 20 days . This method already described [11] was slightly improved as follows . Briefly , Nup133-GFP cells , expressing the WT or mutated BRCA1 protein fused to mCherry , were induced for 4 hours with galactose before analysis using live fluorescent microscopy . The previously named "yeast localization phenotype" ( YLP ) assay [11] was subdivided into two assays in this study . The Spot Formation assay monitors the proportion of cells showing a single aggregate of WT or mutated BRCA1 , visible in fluorescent microscopy , without considering the intracellular localization . This aggregate is also referred to as "spot" . Cells with several aggregates were not considered in this assay . The Yeast Localization assay monitors the proportion of spot volumes localized in the cytoplasm of yeast cells . Picture acquisitions were previously described [11] . For each clone , at least three fields , containing at least 100 cells , were acquired . For the Spot Formation assay , the number of cells showing one spot was manually counted . Next , the proportion of cells containing one spot was computed by dividing the number of cells showing one spot to the total number of cells ( one value per clone ) . For the Yeast Localization assay , images of the three fields were deconvoluted [23] and the volume Volij of each spot i , in the field j , was measured using the 3D Object Counter plugin [24] of ImageJ . Next , each spot was manually categorized as "inside" or "outside" the nucleus . Finally , the proportion of volume outside the nucleus was computed using the formula ( ΣiΣjVolij/outside ) / ( ΣiΣjVolij/outside + ΣiΣjVolij/inside ) , which led to one value per each clone assessed . This proportion quantifies the cytoplasmic localization of the mCherry protein fused to BRCA1 . For the simultaneous assessment of 10 variants by a single technician , using the Spot Formation and Yeast Localization assays , the time required between the delivery of the intermediate plasmids ( see above ) and the final results is 21 days . IGR-CaP1 epithelial cells , derived from a human prostate primary tumor [25] , were plated in 384-well plates at 750 cells/well , were allowed to adhere overnight and then were transfected with a single siRNA from a siRNA library targeting 406 different genes . siKIF11 , siGL2 and siGOLGA2 were used as controls . After 72 hours , cells were fixed and nuclei were stained with DAPI . Images were acquired with an INCell 2000 automated wide-field system ( GE Healthcare , Little Chalfont , UK ) and cell counts were quantified in each well with the INCell Analyzer workstation software ( GE Healthcare ) . The pictures analyzed represent 20% of the well surface , which corresponds to an average of 150 cells initially plated for this surface . Statistical and computational methods , as well as R source codes , are provided in the S1 Text . | Human genetics has entered a new age with the advent of next generation sequencing . However , this great advance also comes with new concerns . Currently , the extensive use of multi-gene panels , whole exome and whole genome sequencing , is generating an ever-growing number of new DNA sequence variations detected in the disease-predisposing genes of human patients . The pathogenic or neutral status of these variants needs to be known before planning any medical act or follow-up . We show here that the status of the variants identified in the BRCA1 breast/ovarian cancer susceptibility gene can be estimated thanks to experimental systems using yeast cells and a novel computational model . Importantly , this model provides a probabilistic classification of variants , opening the possibility to integrate results from functional assays into clinical decision-making . Moreover , our computational model is directly compatible with all kinds of experimental system without any requirement for skills in statistics thanks to ready-to-use online tools . We believe that this work is a step forward in the clinical interpretation of human genetic variants . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"decision",
"making",
"gene",
"regulation",
"pathogens",
"microbiology",
"cloning",
"neuroscience",
"plasmid",
"construction",
"mutation",
"probability",
"distribution",
"mathematics",
... | 2016 | Functional Assessment of Genetic Variants with Outcomes Adapted to Clinical Decision-Making |
Escherichia coli ( E . coli ) bacteria govern their trajectories by switching between running and tumbling modes as a function of the nutrient concentration they experienced in the past . At short time one observes a drift of the bacterial population , while at long time one observes accumulation in high-nutrient regions . Recent work has viewed chemotaxis as a compromise between drift toward favorable regions and accumulation in favorable regions . A number of earlier studies assume that a bacterium resets its memory at tumbles – a fact not borne out by experiment – and make use of approximate coarse-grained descriptions . Here , we revisit the problem of chemotaxis without resorting to any memory resets . We find that when bacteria respond to the environment in a non-adaptive manner , chemotaxis is generally dominated by diffusion , whereas when bacteria respond in an adaptive manner , chemotaxis is dominated by a bias in the motion . In the adaptive case , favorable drift occurs together with favorable accumulation . We derive our results from detailed simulations and a variety of analytical arguments . In particular , we introduce a new coarse-grained description of chemotaxis as biased diffusion , and we discuss the way it departs from older coarse-grained descriptions .
The bacterium E . coli moves by switching between two types of motions , termed ‘run’ and ‘tumble’ [1] . Each results from a distinct movement of the flagella . During a run , flagella motors rotate counter-clockwise ( when looking at the bacteria from the back ) , inducing an almost constant forward velocity of about , along a near-straight line . In an environment with uniform nutrient concentration , run durations are distributed exponentially with a mean value of about [2] . When motors turn clockwise , the bacterium undergoes a tumble , during which , to a good approximation , it does not translate but instead changes its direction randomly . In a uniform nutrient-concentration profile , the tumble duration is also distributed exponentially but with a much shorter mean value of about [3] . When the nutrient ( or , more generally , chemoattractant ) concentration varies in space , bacteria tend to accumulate in regions of high concentration ( or , equivalently , the bacteria can also be repelled by chemorepellants and tend to accumulate in low chemical concentration ) [4] . This is achieved through a modulation of the run durations . The biochemical pathway that controls flagella dynamics is well understood [1] , [5]–[7] and the stochastic ‘algorithm’ which governs the behavior of a single motor is experimentally measured . The latter is routinely used as a model for the motion of a bacteria with many motors [1] , [8]–[11] . This algorithm represents the motion of the bacterium as a non-Markovian random walker whose stochastic run durations are modulated via a memory kernel , shown in Fig . 1 . Loosely speaking , the kernel compares the nutrient concentration experienced in the recent past with that experienced in the more distant past . If the difference is positive , the run duration is extended; if it is negative , the run duration is shortened . In a complex medium bacterial navigation involves further complications; for example , interactions among the bacteria , and degradations or other dynamical variations in the chemical environment . These often give rise to interesting collective behavior such as pattern formation [12] , [13] . However , in an attempt to understand collective behavior , it is imperative to first have at hand a clear picture of the behavior of a single bacterium in an inhomogeneous chemical environment . We are concerned with this narrower question in the present work . Recent theoretical studies of single-bacterium behavior have shown that a simple connection between the stochastic algorithm of motion and the average chemotactic response is far from obvious [8]–[11] . In particular , it appeared that favorable chemotactic drift could not be reconciled with favorable accumulation at long times , and chemotaxis was viewed as resulting from a compromise between the two [11] . The optimal nature of this compromise in bacterial chemotaxis was examined in Ref . [10] . In various approximations , while the negative part of the response kernel was key to favorable accumulation in the steady state , it suppressed the drift velocity . Conversely , the positive part of the response kernel enhanced the drift velocity but reduced the magnitude of the chemotactic response in the steady state . Here , we carry out a detailed study of the chemotactic behavior of a single bacterium in one dimension . We find that , for an ‘adaptive’ response kernel ( i . e . , when the positive and negative parts of the response kernel have equal weight such that the total area under the curve vanishes ) , there is no incompatibility between a strong steady-state chemotaxis and a large drift velocity . A strong steady-state chemotaxis occurs when the positive peak of the response kernel occurs at a time much smaller than and the negative peak at a time much larger than , in line with experimental observation . Moreover , we obtain that the drift velocity is also large in this case . For a general ‘non-adaptive’ response kernel ( i . e . , when the area under the response kernel curve is non-vanishing ) , however , we find that a large drift velocity indeed opposes chemotaxis . Our calculations show that , in this case , a position-dependent diffusivity is responsible for chemotactic accumulation . In order to explain our numerical results , we propose a simple coarse-grained model which describes the bacterium as a biased random walker with a drift velocity and diffusivity , both of which are , in general , position-dependent . This simple model yields good agreement with results of detailed simulations . We emphasize that our model is distinct from existing coarse-grained descriptions of E . coli chemotaxis [13]–[16] . In these , coarse-graining was performed over left- and right-moving bacteria separately , after which the two resulting coarse-grained quantities were then added to obtain an equation for the total coarse-grained density . We point out why such approaches can fail and discuss the differences between earlier models and the present coarse-grained model .
Following earlier studies of chemotaxis [9] , [17] , we model the navigational behavior of a bacterium by a stochastic law of motion with Poissonian run durations . A switch from run to tumble occurs during the small time interval between and with a probability ( 1 ) Here , and is a functional of the chemical concentration , , experienced by the bacterium at times . In shallow nutrient gradients , the functional can be written as ( 2 ) The response kernel , , encodes the action of the biochemical machinery that processes input signals from the environment . Measurements of the change in the rotational bias of a flagellar motor in wild-type bacteria , in response to instantaneous chemoattractant pulses were reported in Refs . [17] , [18]; experiments were carried out with a tethering assay . The response kernel obtained from these measurements has a bimodal shape , with a positive peak around and a negative peak around ( see Fig . 1 ) . The negative lobe is shallower than the positive one and extends up to , beyond which it vanishes . The total area under the response curve is close to zero . As in other studies of E . coli chemotaxis , we take this response kernel to describe the modulation of run duration of swimming bacteria [8]–[11] . Recent experiments suggest that tumble durations are not modulated by the chemical environment and that as long as tumbles last long enough to allow for the reorientation of the cell , bacteria can perform chemotaxis successfully [19] , [20] . The model defined by Eqs . 1 and 2 is linear . Early experiments pointed to a non-linear , in effect a threshold-linear , behavior of a bacterium in response to chemotactic inputs [17] , [18] . In these studies , a bacterium modulated its motion in response to a positive chemoattractant gradient , but not to a negative one . In the language of present model , such a threshold-linear response entails replacing the functional defined in Eq . 2 by zero whenever the integral is negative . More recent experiments suggest a different picture , in which a non-linear response is expected only for a strong input signal whereas the response to weak chemoattractant gradient is well described by a linear relation [21] . Here , we present an analysis of the linear model . For the sake of completeness , in Text S1 , we present a discussion of models which include tumble modulations and a non-linear response kernel . Although recent experiments have ruled out the existence of both these effects in E . coli chemotaxis , in general such effects can be relevant to other systems with similar forms of the response function . The shape of the response function hints to a simple mechanism for the bacterium to reach regions with high nutrient concentration . The bilobe kernel measures a temporal gradient of the nutrient concentration . According to Eq . 1 , if the gradient is positive , runs are extended; if it is negative , runs are unmodulated . However , recent literature [8] , [9] , [11] has pointed out that the connection between this simple picture and a detailed quantitative analysis is tenuous . For example , de Gennes used Eqs . 1 to calculate the chemotactic drift velocity of bacteria [8] . He found that a singular kernel , , where is a Dirac function and a positive constant , lead to a mean velocity in the direction of increasing nutrient concentration even when bacteria are memoryless ( ) . Moreover , any addition of a negative contribution to the response kernel , as seen in experiments ( see Fig . 1 ) , lowered the drift velocity . Other studies considered the steady-state density profile of bacteria in a container with closed walls , both in an approximation in which correlations between run durations and probability density were ignored [11] and in an approximation in which the memory of the bacterium was reset at run-to-tumble switches [9] . Both these studies found that , in the steady state , a negative contribution to the response function was mandatory for bacteria to accumulate in regions of high nutrient concentration . These results seem to imply that the joint requirement of favorable transient drift and steady-state accumulation is problematic . The paradox was further complicated by the observation [9] that the steady-state single-bacterium probability density was sensitive to the precise shape of the kernel: when the negative part of the kernel was located far beyond it had little influence on the steady-state distribution [11] . In fact , for kernels similar to the experimental one , model bacteria accumulated in regions with low nutrient concentration in the steady state [9] .
In order to resolve these paradoxes and to better understand the mechanism that leads to favorable accumulation of bacteria , we perform careful numerical studies of bacterial motion in one dimension . In conformity with experimental observations [17] , [18] , we do not make any assumption of memory reset at run-to-tumble switches . We model a bacterium as a one-dimensional non-Markovian random walker . The walker can move either to the left or to the right with a fixed speed , , or it can tumble at a given position before initiating a new run . In the main paper , we present results only for the case of instantaneous tumbling with , while results for non-vanishing are discussed in Text S1 . There , we verify that for an adaptive response kernel does not have any effect on the steady-state density profile . For a non-adaptive response kernel , the correction in the steady-state slope due to finite is small and proportional to . The run durations are Poissonian and the tumble probability is given by Eq . 1 . The probability to change the run direction after a tumble is assumed to have a fixed value , , which we treat as a parameter . The specific choice of the value of does not affect our broad conclusions . We find that , as long as , only certain detailed quantitative aspects of our numerical results depend on . ( See Text S1 for details on this point . ) We assume that bacteria are in a box of size with reflecting walls and that they do not interact among each other . We focus on the steady-state behavior of a population . Reflecting boundary conditions are a simplification of the actual behavior [22] , [23]; as long as the total ‘probability current’ ( see discussion below ) in the steady state vanishes , our results remain valid even if the walls are not reflecting . As a way to probe chemotactic accumulation , we consider a linear concentration profile of nutrient: . We work in a weak gradient limit , i . e . , the value of is chosen to be sufficiently small to allow for a linear response . Throughout , we use in our numerics . From the linearity of the problem , results for a different attractant gradient , , can be obtained from our results through a scaling factor . In the linear reigme , we obtain a spatially linear steady-state distribution of individual bacterium positions , or , equivalently , a linear density profile of a bacterial population . Its slope , which we denote by , is a measure of the strength of chemotaxis . A large slope indicates strong bacterial preference for regions with higher nutrient concentration . Conversely , a vanishing slope implies that bacteria are insensitive to the gradient of nutrient concentration and are equally likely to be anywhere along the line . We would like to understand the way in which the slope depends on the different time scales present in the system . In order to gain insight into our numerical results , we developed a simple coarse-grained model of chemotaxis . For the sake of simplicity , we first present the model for a non-adaptive , singular response kernel , , and , subsequently , we generalize the model to adaptive response kernels by making use of linear superposition . The memory trace embodied by the response kernel induces temporal correlations in the trajectory of the bacterium . However , if we consider the coarse-grained motion of the bacterium over a spatial scale that exceeds the typical run stretch and a temporal scale that exceeds the typical run duration , then we can assume that it behaves as a Markovian random walker with drift velocity and diffusivity . Since the steady-state probability distribution , , is flat for , for small we can write ( 4 ) ( 5 ) ( 6 ) Here , and . Since we are neglecting all higher order corrections in , our analysis is valid only when is sufficiently small . In particular , even when , we assume that the inequality is still satisfied . The chemotactic drift velocity , , vanishes if ; it is defined as the mean displacement per unit time of a bacterium starting a new run at a given location . Clearly , even in the steady state when the current , defined through , vanishes , may be non-vanishing ( see Eq . 8 below ) . In general , the non-Markovian dynamics make dependent on the initial conditions . However , in the steady state this dependence is lost and can be calculated , for example , by performing a weighted average over the probability of histories of a bacterium . This is the quantity that is of interest to us . An earlier calculation by de Gennes showed that , if the memory preceding the last tumble is ignored , then for a linear profile of nutrient concentration the drift velocity is independent of position and takes the form [8] . While the calculation applies strictly in a regime with ( because of memory erasure ) , in fact its result captures the behavior well over a wide range of parameters ( see Fig . 4 ) . To measure in our simulations , we compute the average displacement of the bacterium between two successive tumbles in the steady state , and we extract therefrom the drift velocity . ( For details of the derivation , see Text S1 . ) We find that is negative for and that its magnitude falls off with increasing values of ( Fig . 4 ) . We also verify that indeed does not show any spatial dependence ( data shown in Fig . of Text S1 ) . We recall that , in our numerical analysis , we have used a small value of ; this results in a low value of . We show below that for an experimentally measured bilobe response kernel , obtained by superposition of singular response kernels , the magnitude of becomes larger and comparable with experimental values . To obtain the diffusivity , , we first calculate the effective mean free path in the coarse-grained model . The tumbling frequency of a bacterium is and depends on the details of its past trajectory . In the coarse-grained model , we replace the quantity by an average over all the trajectories within the spatial resolution of the coarse-graining . Equivalently , in a population of non-interacting bacteria , the average is taken over all the bacteria contained inside a blob , and , hence , denotes the position of the center of mass of the blob at a time in the past . As mentioned above , the drift velocity is proportional to , so that . The average tumbling frequency then becomes and , consequently , the mean free path becomes . As a result , the diffusivity is expressed as . We checked this form against our numerical results ( Fig . 5 ) . Having evaluated the drift velocity , , and the diffusivity , , we now proceed to write down the continuity equation ( for a more rigorous but less intuitive approach , see [10] ) . For a biased random walker on a lattice , with position-dependent hopping rates and towards the right and the left , respectively , one has and , where is the lattice constant . In the continuum limit , the temporal evolution of the probability density is given by a probability current , as ( 7 ) where the current takes the form ( 8 ) For reflecting boundary condition , in the steady state . This constraint yields a steady-state slope ( 9 ) for small . We use our measured values for and ( Figs . 4 and 5 ) , and compute the slope using Eq . 9 . ( For details of the measurement of , see Text S1 . ) We compare our analytical and numerical results in Fig . 2 , which exhibits close agreement . According to Eq . 9 , steady-state chemotaxis results from a competition between drift motion and diffusion . For , the drift motion is directed toward regions with a lower nutrient concentration and hence opposes chemotaxis . Diffusion is spatially dependent and becomes small for large nutrient concentrations ( again for ) , thus increasing the effective residence time of the bacteria in favorable regions . For large values of , the drift velocity vanishes and one has a strong chemotaxis as increases ( Fig . 2 ) . Finally , for , the calculation by de Gennes yields which exactly cancels the spatial gradient of ( to linear order in ) , and there is no accumulation [8] , [11] . These conclusions are easily generalized to adaptive response functions . For , within the linear response regime , the effective drift velocity and diffusivity can be constructed by simple linear superposition: The drift velocity reads . Interestingly , the spatial dependence of cancels out and . The resulting slope then depends on the drift only and is calculated as ( 10 ) In this case , the coarse-grained model is a simple biased random walker with constant diffusivity . For and , the net velocity , proportional to , is positive and gives rise to a favorable chemotactic response , according to which bacteria accumulate in regions with high food concentration . Moreover , the slope increases as the separation between and grows . We emphasize that there is no incompatibility between strong steady-state chemotaxis and large drift velocity . In fact , in the case of an adaptive response function , strong chemotaxis occurs only when the drift velocity is large . For a bilobe response kernel , approximated by a superposition of many delta functions ( Fig . 1 ) , the slope , , can be calculated similarly and in Fig . 3 we compare our calculation to the simulation results . We find close agreement in the case of a linear model with a bilobe response kernel and , in fact , also in the case of a non-linear model ( see Text S1 ) . The experimental bilobe response kernel is a smooth function , rather than a finite sum of singular kernels over a set of discrete values ( as in Fig . 1 ) . Formally , we integrate singular kernels over a continuous range of to obtain a smooth response kernel . If we then integrate the expression for the drift velocity obtained by de Gennes , according to this procedure , we find an overall drift velocity , for the concentration gradient considered ( ) . By scaling up the concentration gradient by a factor of , the value of can also be scaled up by and can easily account for the experimentally measured velocity range .
We carried out a detailed analysis of steady-state bacterial chemotaxis in one dimension . The chemotactic performance in the case of a linear concentration profile of the chemoattractant , , was measured as the slope of the bacterium probability density profile in the steady state . For a singular impulse response kernel , , the slope was a scaling function of , which vanished at the origin , increased monotonically , and saturated at large argument . To understand these results we proposed a simple coarse-grained model in which bacterial motion was described as a biased random walk with drift velocity , , and diffusivity , . We found that for small enough values of , was independent of and varied linearly with nutrient concentration . By contrast , was spatially uniform and its value decreased monotonically with and vanished for . We presented a simple formula for the steady-state slope in terms of and . The prediction of our coarse-grained model agreed closely with our numerical results . Our description is valid when is small enough , and all our results are derived to linear order in . We assume is always satisfied . Our results for an impulse response kernel can be easily generalized to the case of response kernels with arbitrary shapes in the linear model . For an adaptive response kernel , the spatial dependence of the diffusivity , , cancels out but a positive drift velocity , , ensures bacterial accumulation in regions with high nutrient concentration , in the steady state . In this case , the slope is directly proportional to the drift velocity . As the delay between the positive and negative peaks of the response kernel grows , the velocity increases , with consequent stronger chemotaxis . Earlier studies of chemotaxis [13]–[16] put forth a coarse-grained model different from ours . In the model first proposed by Schnitzer for a single chemotactic bacterium [14] , he argued that , in order to obtain favorable bacterial accumulation , tumbling rate and ballistic speed of a bacterium must both depend on the direction of its motion . In his case , the continuity equation reads ( 11 ) where is the ballistic speed and is the tumbling frequency of a bacterium moving toward the left ( right ) . For E . coli , as discussed above , , a constant independent of the location . In that case , Eq . 11 predicts that in order to have a chemotactic response in the steady state , one must have a non-vanishing drift velocity , i . e . , . This contradicts our findings for non-adaptive response kernels , according to which a drift velocity only hinders the chemotactic response . The spatial variation of the diffusivity , instead , causes the chemotactic accumulation . This is not captured by Eq . 11 . In the case of adaptive response kernels , the diffusivity becomes uniform while the drift velocity is positive , favoring chemotaxis . Comparing the expression of the flux , , obtained from Eqs . 7 and 8 with that from Eq . 11 , and matching the respective coefficients of and , we find and . As we argued above in discussing the coarse-grained model for adaptive response kernels , both and are spatially independent . This puts strict restrictions on the spatial dependence of and . For example , as in E . coli chemotaxis , our coarse-grained description is recovered only if and are also independent of . We comment on a possible origin of the discrepancy between our work and earlier treatments . In Ref . [14] , a continuity equation was derived for the coarse-grained probability density of a bacterium , starting from a pair of approximate master equations for the probability density of a right-mover and a left-mover , respectively . As the original process is non-Markovian , one can expect a master equation approach to be valid only at scales that exceed the scale over which spatiotemporal correlations in the behavior of the bacterium are significant . In particular , a biased diffusion model can be viewed as legitimate only if the ( coarse-grained ) temporal resolution allows for multiple runs and tumbles . If so , at the resolution of the coarse-grained model , left- and right-movers become entangled , and it is not possible to perform a coarse-graining procedure on the two species separately . Thus one cannot define probability densities for a left- and a right-mover that evolves in a Markovian fashion . In our case , left- and right-movers are coarse-grained simultaneously , and the total probability density is Markovian . Thus , our diffusion model differs from that of Ref . [14] because it results from a different coarse-graining procedure . The model proposed in Ref . [14] has been used extensively to investigate collective behaviors of E . coli bacteria such as pattern formation [13] , [15] , [16] . It would be worth asking whether the new coarse-grained description can shed new light on bacterial collective behavior . | The chemotaxis of Escherichia coli is a prototypical model of navigational strategy . The bacterium maneuvers by switching between near-straight motion , termed runs , and tumbles which reorient its direction . To reach regions of high nutrient concentration , the run-durations are modulated according to the nutrient concentration experienced in recent past . This navigational strategy is quite general , in that the mathematical description of these modulations also accounts for the active motility of C . elegans and for thermotaxis in Escherichia coli . Recent studies have pointed to a possible incompatibility between reaching regions of high nutrient concentration quickly and staying there at long times . We use numerical investigations and analytical arguments to reexamine navigational strategy in bacteria . We show that , by accounting properly for the full memory of the bacterium , this paradox is resolved . Our work clarifies the mechanism that underlies chemotaxis and indicates that chemotactic navigation in wild-type bacteria is controlled by drift while in some mutant bacteria it is controlled by a modulation of the diffusion . We also propose a new set of effective , large-scale equations which describe bacterial chemotactic navigation . Our description is significantly different from previous ones , as it results from a conceptually different coarse-graining procedure . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] | [
"physics",
"statistical",
"mechanics",
"theoretical",
"biology",
"biophysics",
"theory",
"biology",
"computational",
"biology",
"biophysics",
"simulations",
"biophysics"
] | 2011 | Chemotaxis when Bacteria Remember: Drift versus Diffusion |
Gene expression signatures that are predictive of therapeutic response or prognosis are increasingly useful in clinical care; however , mechanistic ( and intuitive ) interpretation of expression arrays remains an unmet challenge . Additionally , there is surprisingly little gene overlap among distinct clinically validated expression signatures . These “causality challenges” hinder the adoption of signatures as compared to functionally well-characterized single gene biomarkers . To increase the utility of multi-gene signatures in survival studies , we developed a novel approach to generate “personal mechanism signatures” of molecular pathways and functions from gene expression arrays . FAIME , the Functional Analysis of Individual Microarray Expression , computes mechanism scores using rank-weighted gene expression of an individual sample . By comparing head and neck squamous cell carcinoma ( HNSCC ) samples with non-tumor control tissues , the precision and recall of deregulated FAIME-derived mechanisms of pathways and molecular functions are comparable to those produced by conventional cohort-wide methods ( e . g . GSEA ) . The overlap of “Oncogenic FAIME Features of HNSCC” ( statistically significant and differentially regulated FAIME-derived genesets representing GO functions or KEGG pathways derived from HNSCC tissue ) among three distinct HNSCC datasets ( pathways:46% , p<0 . 001 ) is more significant than the gene overlap ( genes:4% ) . These Oncogenic FAIME Features of HNSCC can accurately discriminate tumors from control tissues in two additional HNSCC datasets ( n = 35 and 91 , F-accuracy = 100% and 97% , empirical p<0 . 001 , area under the receiver operating characteristic curves = 99% and 92% ) , and stratify recurrence-free survival in patients from two independent studies ( p = 0 . 0018 and p = 0 . 032 , log-rank ) . Previous approaches depending on group assignment of individual samples before selecting features or learning a classifier are limited by design to discrete-class prediction . In contrast , FAIME calculates mechanism profiles for individual patients without requiring group assignment in validation sets . FAIME is more amenable for clinical deployment since it translates the gene-level measurements of each given sample into pathways and molecular function profiles that can be applied to analyze continuous phenotypes in clinical outcome studies ( e . g . survival time , tumor volume ) .
The application of gene signatures to clinical outcome prediction has become an area of intensive research . In cancer , expression signatures of poor prognosis [1] , recurrence [2] , invasiveness [3] , metastasis [4] , and therapeutic response [5] , [6] have been developed using either data-driven approaches in clinical trials , or via biologically validated mechanisms found prior to the clinical trials . However , gene lists of distinct signatures do not significantly overlap [7] , [8] , even though they paradoxically occupy a common prognostic space and are similarly efficient in predicting poor clinical outcomes in new cohorts . These observations have raised questions about their biologic relevance , significance and clinical implication [7] , [8] . New types of mechanism-anchored gene expression signatures are highly desirable for personal genomics but are currently unavailable for single sample prognosis of continuous quantitative phenotypes ( e . g . survival time ) . Since commercial microarrays are now a mature commercial technology and could become a reliable data source amenable to clinical practice , we were motivated to investigate the remaining barriers to their applications in personal genomics . Aside from differences in computational methods used for deriving gene expression signatures , several hypotheses have been postulated to explain the lack of gene overlap and low reproducibility of the genetic makeup among existing expression signatures . One explanation is that different genes are merely separate aspects of the same groups of molecular pathways or mechanisms [8] , [9] . This hypothesis has been examined using the Kyoto Encyclopedia of Genes and Genomes ( KEGG ) [10] or Gene Ontology ( GO ) [11] to derive functionally related gene-sets as mechanism-anchored signatures from microarray profiling [12] , or from a priori knowledge and experimental genome-wide expression data [13] . For conducting such analyses , various analytical and statistical methods have been developed such as DAVID [14] , GOstat [15] , FunCluster [16] , FunNet [17] , GSEA [18] , MGSA [19] , principal component analysis [20] , FatiScan [21] and globaltest [22] . These conventional methods of functional gene-set analysis ( reviewed in [23] , [24] ) have improved our overall ability for identifying dysregulated mechanisms from gene expression of a cohort of patients [20] , [25]–[29] , however they cannot , by design , provide pathway scores at the single sample level . Thus , their potential for clinical usage is limited . Developing the capacity to provide an individualized mechanistic interpretation of analysis results as they relate to clinical outcomes or treatment strategies , will greatly enhance the clinical deployment of signatures . The state-of-art data-driven but rate limiting methods for generating pathway signatures focus on the coordinated changes in expression of multiple genes in a pathway experimentally detected in animal models [30] or on the knock-in or -down of a key pathway gene in human cells [31] , [32] . Recently , two types of knowledge-driven approaches have also been proposed for generating pathway signatures directly from human tumor specimens [33] ( a ) those using the straightforward unsupervised pathway measures ( e . g . , mean , median expression of all pathway gene members ) within each sample [28] , [34] , and ( b ) those generating pathway scores after performing supervised statistics requiring sample class assignment ( e . g . principal component analysis , PCA [35]–[37] , CORG “condition-responsive genes” [27] , LLR [38] ) . While the latter set of methods is more accurate [27] , the dependencies between samples preclude their utility for single-sample prognostication . Furthermore , pathway signatures derived from these state-of-the-art methods have been validated in predicting qualitative clinical outcomes , such as complete remission vs . disease progression . These methods , however , are not designed for making prediction using continuous clinical measures , such as recurrence-free survival time [27] , [38] . Therefore , novel bioinformatics approaches are required for single-sample assignment of biological features from gene expression analyses so that the wealth of seemingly uninterpretable molecular data can be translated into mechanistic interpretations , which can in turn be utilized for making therapeutic choices and forecasting clinical outcomes . We hypothesized that molecular mechanisms delineated from gene expression deregulation profiles are accessible as genome-wide measurements of pathways at a single-sample level . Here , we present a novel methodology , the Functional Analysis of Individual Microarray Expression ( FAIME ) that can translate patient microarray data into pathway and molecular functional profiles on a single-sample level and can be applied to quantitative phenotypes of outcome prediction ( e . g . survival time , tumor volume response to therapy ) . FAIME , by computing statistical scores on individual patients , retains sample independence within a cohort and enables subsequent mechanism-level clustering or signature validation . We demonstrate the potential of FAIME in personalized genomics using relatively small-size cohorts of Head and Neck Squamous Cell Cancer ( HNSCC ) in which FAIME produces single-patient survival prediction . Ninety percent of patients with HNSCC will present with disease that is locoregionally confined and will be considered for curative intent therapy [39] . However , individual outcome prognostication is poor because it is based almost entirely on tumor anatomic location and size [40] . Presently , all patients who are candidates for curative intent treatment are offered a multimodality approach that is associated with serious acute toxicity and long-term dysfunction [41] since there are no reliable indicators to predict response to therapy . Treatment usually consists of broadly cytotoxic entities ( e . g . radiation , chemotherapy ) , and pathobiology based targeted therapies are few [42] . Not surprisingly , we have shown a strong correlation between response to induction therapy and survival [42] . Nevertheless , there are currently no validated pre-treatment classifiers to discriminate the fraction of patients that will benefit clinically from those who will not . Therefore , accurate mechanistic derived signatures would provide valuable prognostic information , the ability to select patients for appropriately intense treatment , and potentially help identify novel targets that could be integrated into current therapy .
The FAIME method provides a translation of each sample's gene expression to molecular mechanisms ( FAIME-score ) . We conduct three analyses to compare the robustness of FAIME to that of previous approaches . We first examine the stability of FAIME Scores within samples , followed by demonstration of the reproducibility of deregulated FAIME-derived mechanisms across three independent HNSCC datasets , and the determination of the precision and recall of FAIME mechanisms using a “proxy gold standard” ( Methods ) as a measure of concordance between FAIME predictions and other methods . In addition to its ability to discriminate between the non-tumor control and HNSCC tumor tissues , the 57 Oncogenic FAIME Features of HNSCC , as a whole , have significantly higher prognostic power . They stratify HNSCC samples in two distinct HNSCC datasets ( E and F ) into two recurrence-free survival subgroups ( log-rank p = 0 . 0018 and 0 . 032 , dataset E and F , respectively; Figure 6 ) . This survival analysis exemplifies a distinctive task that FAIME was designed to accomplish: sample's mechanisms scores are calculated without group assignment and thus circumvent the risk of overtraining . We further validate the prognostic power of FAIME to predict survival of patients from an additional HNSCC dataset previously not used in this study , GSE2837 ( Log-rank p = 0 . 049 , Kaplan-Meier curve in Figure S5 ) . We then compare the utility of FAIME-derived scores with other enrichment methods that also do not require phenotypic group assignment . In contrast , Enrichment analysis , Mean-G and GSEA scores fail to provide stratification of patients by RFS time ( p>27% , p>7% , and p>34%; smallest log-rank p reported from dataset E or F; Enrichment , Mean-G and GSEA respectively; Methods ) . These results attest to the utility of FAIME for learning predictive mechanism patterns from gene expression in pursuance of quantitative phenotypes such as survival analysis . In order to compare the prognostic power of FAIME to other prognosticators of HNSCC , we note that for dataset E in this study , the authors report a non-statistically significant trend for HPV+ HNC patients to have better survival outcomes when compared to HPV- HNC patients ( HNSCCs classified according to Pyeon HPV geneset [46]: p = 0 . 09 and p = 0 . 33 for two independent datasets; HNSCCs classified HPV+/− by p16 IHC: p = 0 . 20 and p = 0 . 52 for two independent datasets ) . In contrast , we show that when FAIME is run on this dataset , we obtain a significant prognostic indicator for survival outcome for datasets E , F , and GSE2837 ( p = 0 . 0018 , p = 0 . 032 , p = 0 . 049 , respectively ) in Figure 6 and Figure S5 . We next investigate if these FAIME-derived patterns could recapitulate known biological and pathophysiological knowledge . As shown in Figure S6 , six of the 57 Oncogenic FAIME Features of HNSCC significantly overlap with the KEGG pathways and GO-MF that can be derived by enrichment of the 31 genes ( 54 probes ) associated with disease-specific survival in a study by Thurlow et al . [47] . In the second study by Perou and colleagues , a set of 582 deregulated genes has classified HNSCC into four “intrinsic” groups ( I–IV ) , of which some combinations are associated with poorer recurrence-free survival [48] ( dataset F , Table 1 ) . Importantly , FAIME can recapitulate at the mechanism level this molecular classification: by producing four groups using an unsupervised method , the 57 Oncogenic FAIME Features of HNSCC are enriched in the four original “intrinsic” groups of Perou's molecular classification of HNSCC samples ( p = 0 . 0031 , Methods , Dataset F , Fisher's Exact Test on a 4×4 contingency table ) . It has been demonstrated that some multi-gene expression signature classifiers , derived from comparisons made between the cancer and control tissues , can provide both diagnostic stratification of clinical tumor samples as expected , as well as prognostic prediction of clinical outcomes , such as RFS or overall survival across patients within a clinical cohort [47] , [48] . However , to our knowledge this is the first mechanism-level predictor , generated from gene expression changes between the cancer and control tissues , that possess both diagnostic and prognostic power at the level of each individual clinical sample without requiring group assignment in validation sets . These observations also indicate that common molecular mechanisms may underlie oncogenesis and disease recurrence . Therefore , therapeutic targeting of such common mechanisms may have potential clinical benefits of effective local control of primary tumors , and at the same time , preventing disease recurrence . In order to prioritize other potential RFS mechanisms and to further demonstrate the utility of FAIME Scores for the study of quantitative phenotypes ( e . g . RFS ) , we conduct a Cox regression analysis of recurrence-free survival time for each of the 208 KEGG pathways and 956 GO terms for which a FAIME Score could be computed in datasets E and F with four or more genes per mechanism ( Methods , Table 1 ) . Two significant mechanism genesets can each be considered as a “single mechanism” prognostic predictor of RFS for future clinical validation: ( i ) [hsa04210] , and ( ii ) receptor signaling complex scaffold activity [GO:0030159] ( p = 0 . 0026 and 0 . 0034 respectively; Bonferroni-adjusted meta-analysis across Datasets E and F of the Cox regressions Table S3 , Methods ) . Since disease recurrence is also a predictor of RFS , FAIME Scores of these pre-treatment samples are also significantly decreased in patients with rapid onset of disease recurrence in both datasets E and F ( Table S3 , Mann-Whitney test ) . As a proof-of-concept study we focus on the topmost candidate RFS mechanism , and conduct a Principal Component Analysis ( PCA ) of datasets E and F for the prioritized KEGG apoptosis geneset ( KEGG pathway [hsa04210] containing known genes associated with apoptosis pathways annotated in humans ) since PCA can provide an unbiased approach to derive a metric representing the highest biological variation across samples that can be associated with the biological mechanism ( s ) responsible for this variation . As expected , PCA's first component of the KEGG apoptosis geneset gene expression is also a predictor of deregulated RFS ( Cox regression p = 0 . 0027 and 0 . 044 , Datasets E and F ) . Furthermore , this first component and the FAIME Scores of the KEGG apoptosis geneset are correlated ( Spearman p = 1 . 3×10−8 and 0 . 047 in datasets E and F respectively , Table S3 ) . The detailed list of genes of the KEGG canonical apoptosis geneset is listed in Table S4 . Additionally , we show that FAIME Scores of the KEGG apoptosis geneset are significantly increased in the HNSCC tissues of patients with no evidence of disease recurrence as compared to those of patients with recurrence within both datasets E and F ( p = 0 . 0015 and p = 00051 , respectively , Figure S7 ) . Consistent with these results , HNSCC patients are treated with radiotherapy and conventional chemotherapy regimens that should subsequently induce apoptosis through DNA damage cell cycle checkpoints . Therapy-resistant tumors are more likely to recur early , which is consistent with the observed reduction in the FAIME apoptosis geneset score in patient samples with disease recurrence reflecting a reduced capability of checkpoints to elicit an apoptotic response to therapy-induced DNA damage . The correlation of reduced level of expression of the KEGG apoptosis geneset reported by the FAIME Score in patients with more rapid onset of disease recurrence may reflect aberrant cell cycle checkpoint and DNA damage repair regulation leading to an enhanced survival of HNSCC cancer cells in recurrent patients . This computational FAIME method identifies deregulated genesets associated to mechanisms , and after validation of the FAIME Scores in a prospective study , we will also pursue to validate the deregulation of mechanisms associated to this genesets , such as the apoptosis geneset . Since a pathway-level classifier can be considered as a significant predictor based on the ensemble effect of its constituent gene in a patient , these same genes need not be consistently deregulated in each patient . In other words , the effect is measured at the mechanism level .
FAIME Scores are designed to identify molecular mechanisms whose constituent genes are predominantly up- or down-deregulated , but not both together . We thus regard the present design as a stepping-stone that improves clinical and biological interpretability by reducing the number of features as compared to gene expression signature classifiers , and by obtaining increased statistical power that allows for the inclusion or refutation of each mechanism at the single patient level . Alternatively , the gene expression changes in both directions could be assessed indiscriminately so that subtle changes in molecular mechanisms due to the opposite effect of inhibitor and activator genes of a given pathway can also be identified . Better still , pathway annotations that indicate gene inhibitors and gene enhancers of signals could be pooled together according to the logical direction of the biological significance in the pathway . In principle , FAIME Scores are applicable to other scales of individual quantitative genomic data annotated into genesets from a knowledgebase , for example , from single protein activity measurements . For these extensions , alternative decreasing weights and their effects on FAIME Score need to be discussed . In this manuscript , the weighting strategy used by the FAIME Score is a rank based decreasing from the highest expressed genes to non-expressed genes in individual samples . We chose the weights to decay exponentially based on previous modeling of expression data we conducted [49] , [50] ( Equation 1 ) , arguably future studies may explore alternate models , particularly for different types of genomic data that have not been modeled . Currently , the effort required for interpreting the biological significance of individual genes in expression signature classifiers is challenging . As shown by Bild et al , mechanism-based predictors can bring us one step closer to guiding targeted therapies by using oncogenic pathways derived from cellular experiments [35] . Additional improvements are required to translate FAIME technology for its use in a clinical setting . Specifically , we intend to prospectively validate the capability of Oncogenic FAIME Features of HNSCC to predict survival in a cohort of HNSCC patients treated with a ) cytotoxic chemotherapy ( e . g . induction chemotherapy ) , b ) radiation , and/or c ) an EGFR inhibitor . Future studies will test FAIME's ability to distinguish between patient tumor subtypes: 1 ) HPV positive vs . HPV negative , 2 ) low risk for failure vs . high risk for failure to therapy , and 3 ) emerging genetic drivers of HNSCC ( ongoing TCGA cancer genome atlas effort in head and neck cancer ) . Here , we describe the development and evaluation of FAIME , a computational rather than biological approach designed specifically with the intent for enabling clinicians to functionally interpret individual patient samples using quantitative phenotypes . Additional improvements are required to translate this technology into clinical care , such as the ability to directly interpret a single microarray without cross sample normalization . For example , the “Gene Expression Barcode” relies on a reference standard to interpret single array gene expression [51] , [52] . Furthermore , we and others have shown analytical approaches that do not require a reference standard nor cross-sample normalization to interpret virus genesets at the single pan-microbial array level [53] , [54] whose clinical utility has been documented in some case reports and studies [55] , [56] . In principle , these approaches could be used jointly with FAIME to improve the efficiency of single sample analyses . We note , however , that our analysis of differential gene expression and corresponding mechanisms of head and neck cancer may be complicated by the process of “field cancerization , ” whereby an area of epithelium is preconditioned by a carcinogenic agent , and carcinoma may subsequently arise from multifocal areas [57] . Thus , future expression analyses should scrutinize obtaining non-tumor tissue controls as to avoid the effect of “field cancerization” by using matched controls . Additionally , testing datasets analyzed in this study obtained gene expression data from Affymetrix and Agilent platforms , confirming that our approach is amenable to different gene expression platforms and future studies may utilize additional tissue arrays . Indeed , we recognized that different extraction methods of the tissue , including laser microdissection , could contribute to generating different classifiers . Conceivably , classifiers built from non-laser microdissected data may reveal important genetic signatures encompassing tumor interactions with surrounding stromal cells , but more precise methods are needed to isolate these cells . We note that all tumor samples in the three training datasets and testing datasets D , GSE2837 , and GSE9844 contain at least 70% tumor cells ( Table S5 ) . We provide in Figure S4 showing the validation of the Oncogenic FAIME Features of HNSCC in an independent laser microdissected tumor and non-tumor tissue dataset GSE9844 with a positive predictive value = 92% ( 3 misclassifications ) and p = 0 . 04 ( rank statistics , 100 feature permutations , hierarchical clustering analysis using R ) . Furthermore , in our comparison of laser microdissected tumor tissue with non-tumor control tissue ( GSE9844 ) , we identified 751 out of 2 , 699 significantly deregulated FAIME Scores that were calculated using genesets of mechanisms ( q-value<0 . 05 , z-test adjusted for multiple comparisons ) . Of the 57 Oncogenic FAIME Features of HNSCC ( 57 GO and KEGG genesets ) , 47 were confirmed among these 751 and are deregulated in the same up or down directions originally observed in datasets A , B , and C ( p<10e-16 , enrichment study using Fischer Exact Test , Table S6 ) . Furthermore , we found that Oncogenic FAIME Features of HNSCC were able to accurately classify laser microdissected tumor tissues from non-tumor tissues ( GSE9844 , n = 38 , 3 misclassifications using conventional unsupervised two-way hierarchical clustering ) . We have thus demonstrated that utilizing FAIME Scores to classify laser microdissected tumor tissue is not only feasible , but is also concordant with results from non-laser microdissected tissue datasets in this study . Conventional pathway level classifiers obtained from in vitro/in vivo biological experiments are predetermined and rate limiting for discovering multiple oncogenic pathways or mechanisms , as each require their own biased experimentation [31] , [32] , [35] , [58] . In contrast , in silico knowledge driven approaches are high throughput and can analyze , in an unbiased manner , a larger number of biological mechanisms . Improved reproducibility is achieved with machine-learned mechanism-based predictors that use group assignment ( e . g . CORG [27] , LLR [38] ) , as compared to mechanism predictors derived from the straightforward scoring methods such as Mean-G and Median-G . While “group assignment”-dependent methods are effective in imputing mechanisms for qualitative phenotypes , they are not designed for imputing from a quantitative phenotype . We have designed FAIME to address this challenge . We show that a FAIME-derived predictor composed of Oncogenic FAIME Features of HNSCC can withstand regression analyses of continuous survival time data to predict disease free survival . Beyond predictor construction , we have shown that FAIME-derived mechanisms controlled for multiple comparisons can be retained independently as single mechanism classifiers . We also demonstrate that Oncogenic FAIME Features of HNSCC can classify reproducibly head and neck cancer survival , and have the potential for new knowledge discovery . Furthermore , this approach can conservatively prioritize three classifiers based on a single mechanism classifier for experimental or clinical follow-up validation . Future studies will address FAIME's applicability to a variety of other cancers and tissue-oriented diseases , including oncogenic FAIME Scores of colon cancer metastases progression and of prostate cancer recurrence that are underway . Survival in HNSCC is currently assessed with a mix of TNM staging and biomarker status , such as Human Papilloma Virus ( HPV ) . The utility of gene expression signatures in HNSCC lags behind better suited ones developed in other cancers , such as breast cancer , where expression classifiers are commercially available ( e . g . MammaPrint® microarray [59] . Table S2 shows that the accuracy of Oncogenic FAIME Features of HNSCC to classify tumor vs . non-tumor tissue is comparable to or outperforms those of gene expression classifier signatures in head and neck cancers from the respective papers . However , gene expression signatures are not utilized in clinical settings for head and neck cancers , in part because of their lack of genetic overlap . Deregulated GO and KEGG FAIME Scores have shown a significantly higher overlap between datasets ( Figure 2 ) than deregulated genes between these datasets ( Legend of Figure 2 ) . This reproducibility of FAIME Scores addresses a crucial problem of gene-level expression signatures illustrated by Dr . Joan Massagué in an editorial of the New England Journal of Medicine [60] . In summary , Massagué points out that gene signatures have surprisingly very little overlap when designed in different cohorts , even though they may be equally predictive of the same clinical outcome in these cohorts . In the future , the survival prediction of FAIME Scores is expected to help identify patients , among those for which traditional clinical indicators are inadequate , at high risk of treatment failure that would benefit from more intensive therapy .
Eight gene expression microarray datasets pertaining to HNSCC are used: three for learning expression patterns and for demonstrating concordances of FAIME ( Table 1: datasets A–C [61]–[63] ) , while five other datasets are used for validation ( Table 1: datasets D–F [63] , [47] , [48] , Figure S4 laser-microdissected dataset GSE9844 [64] and Figure S5 dataset GSE2837 [65] ) . The samples of the validation datasets do not overlap with the learning datasets . We define non-tumor control samples as ( i ) samples from an independent , non-smoker individual with no history of HNSCC , ( ii ) paired samples from a distant uninvolved site in patient with HNC ( >3 cm or contralateral ) , and ( iii ) paired samples from the margin of a tumor . For each of these three types of control samples , Table S7 provides the percentage of samples per anatomical location . Several of the non-tumor tissue control samples are enriched specifically for epithelial cells ( datasets B , E , F , GSE2837 , GSE9844 ) , the cell type of origin for head and neck cancer cells , while other non-tumor controls were extracted from mucosae tissues ( datasets A , C and D ) . Additionally , a description of the TNM stage , P53 status , HPV status , smoking status and alcohol intake status is reported in Table S8A–8C . In order to objectively assess the accuracy of the significantly deregulated GO-MF and KEGG pathways identified in each of the three HNSCC datasets identified by FAIME ( Figure 3 ) , a gold standard comprising the true KEGG and GO terms should be used . However , such a gold standard does not exist and published enrichment studies that generated large lists of candidate GO terms and KEGG pathways cannot be thoroughly validated experimentally in their entirety because of the rate limiting nature and cost of such an endeavor . Nonetheless , since a sufficient subset of individual predictions of deregulated molecular functions and pathways from these enrichment studies have been confirmed experimentally , we proceed in using conventional enrichment methods as “proxy-gold standards” . Specifically , GSEA and the hypergeometric enrichments are alternating as proxy-gold standards and as positive controls . These enrichment methods and FAIME are applied to three distinct datasets as described in Stability and Robustness of FAIME Scores with and without cross-sample normalization and gene filtering ( Table 1 , datasets A–C ) . The union of all prioritized GO-MF or KEGG pathways of GSEA or of the hypergeometric analysis illustrated in Figure 3 is alternatively used as a proxy gold standard . Precision and recall is thus calculated for the FAIME results of a specific dataset and that of the remaining conventional method not used to generate the proxy gold standard . The latter serves as a positive conventional control to compare the accuracy of FAIME-derived results . The FAIME-derived mechanism scores of the 208 measurable KEGG pathways and 956 GO-MF terms with more than 3 gene members are calculated for each of the 71 patient samples in dataset E and the 60 patient samples in dataset F . In each dataset , the RFS prognostic power is assessed using a Cox proportional hazards regression with the Bioconductor package survival ( default parameters for censured data ) [81] . In each dataset , a cohort-specific prognostic p-value can thus be calculated for each of these 1 , 164 mechanisms using the Cox-regression analysis ( Bioconductor package survival ) . A meta-analysis of these Cox-regression p-values is then performed using the Stouffer Z-transform method [82] that produces a joint p-value for each mechanism . At each threshold of joint p-values , we obtain a list of prioritized mechanisms for RFS prognosis . Individual FAIME-derived predictors of survival are identified according to a significance threshold of adjusted-p<0 . 05 after controlling the Stouffer meta-analysis for multiple comparisons using the conservative Bonferroni adjustment ( Cox Proportional Hazard applied to datasets E and then F , taking into account the direction of the association for the sign of the Stouffer Z score ) . As a validation , the gene members of significant RFS genes are extracted to conduct the principal component analysis . For each mechanism , the resultant 1st component is compared with FAIME Scores for the RFS prognosis analysis . Bioconductor package amap is employed to run the PCA analysis and the values of the resultant 1st component are compared with FAIME Scores for each patient in dataset E and F . Two types of analyses are conducted: 1 ) the non-parametric Spearman correlation is calculated between 1st component values and FAIME Scores across all samples , and 2 ) the Cox regression analyses of RFS are conducted for both individual FAIME Scores and 1st component values for each sample . Note that for the patient with duplicate measured samples in dataset E , we use the mean expression value to represent its expression measurement and genes with non-measurable expression values are assigned to a value of zero in Dataset F . | Clinical utilization of multi-gene expression signatures that are predictive of therapeutic response has been steadily increasing , however , interpretation of such results remains challenging because multi-gene signatures , generated from analyzing different patient cohorts , tend to be equally predictive but contain minimal overlap . Whereas pathway-level analyses of expression arrays show promise for generating clinically meaningful mechanistic signatures , current approaches do not permit single-patient based analyses that are independent of cross-group calculations . To bridge the gap between deterministic biological mechanisms of single-gene biomarkers and the statistical predictive power of multi-gene signatures that are disconnected from mechanisms , we developed FAIME , a novel method that transforms microarray gene expression data into individualized patient profiles of molecular mechanisms . We have validated its capability for predicting clinical outcomes , including cancer patient samples derived from six different clinical trial cohorts of head and neck cancers . This method provides opportunities to harness an untapped resource for personal genomics: clinical evaluation and testing of individually interpretable mechanistic profiles derived from gene expression arrays . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"oncology",
"medicine",
"genomics",
"biology",
"computational",
"biology",
"cancers",
"and",
"neoplasms",
"genetics",
"and",
"genomics"
] | 2012 | Single Sample Expression-Anchored Mechanisms Predict Survival in Head and Neck Cancer |
Ribosome profiling produces snapshots of the locations of actively translating ribosomes on messenger RNAs . These snapshots can be used to make inferences about translation dynamics . Recent ribosome profiling studies in yeast , however , have reached contradictory conclusions regarding the average translation rate of each codon . Some experiments have used cycloheximide ( CHX ) to stabilize ribosomes before measuring their positions , and these studies all counterintuitively report a weak negative correlation between the translation rate of a codon and the abundance of its cognate tRNA . In contrast , some experiments performed without CHX report strong positive correlations . To explain this contradiction , we identify unexpected patterns in ribosome density downstream of each type of codon in experiments that use CHX . These patterns are evidence that elongation continues to occur in the presence of CHX but with dramatically altered codon-specific elongation rates . The measured positions of ribosomes in these experiments therefore do not reflect the amounts of time ribosomes spend at each position in vivo . These results suggest that conclusions from experiments in yeast using CHX may need reexamination . In particular , we show that in all such experiments , codons decoded by less abundant tRNAs were in fact being translated more slowly before the addition of CHX disrupted these dynamics .
Translation is the process by which the assembly of a protein is directed by the sequence of codons in a messenger RNA . Ribosomes mediate this conversion of information from codons into amino acids through the sequential binding of tRNAs [1] . During the incorporation of each successive amino acid , there are several stages at which the identity of the codon being translated may potentially influence the speed with which a ribosome advances along a coding sequence . When a codon is presented in the A-site of a ribosome , an appropriate tRNA must diffuse into the A-site and successfully form a codon-anticodon base pairing interaction [2 , 3] . tRNAs decoding different codons are expressed at different abundances [4 , 5] , suggesting that ribosomes could spend longer waiting for less abundant tRNAs to arrive [6] . Because translation is accomplished with fewer tRNA identities than there are codon identities , some codon-anticodon interactions involve non-Watson-Crick base-pairings [7 , 8] . These so-called wobble pairings are thought to modulate the speed of decoding [9–12] . Once a tRNA has arrived and base-paired , the speed of peptide bond formation between the C-terminal amino acid in the nascent chain and an incoming amino acid may be influenced by chemical properties of these amino acids [13] . The relative contributions of these effects to overall rates of translation remain poorly understood . Because the genetic code that governs the process of translation maps 61 codon identities to only 20 standard amino acids , multiple synonymous codons can be used to encode most amino acids . There is a rich body of theoretical work on the role of translation speed as a selective force shaping synonymous codon usage [14] , but the ability to directly measure the speed with which each codon is translated in vivo in order to test these theories has historically been lacking . The recent development of ribosome profiling , the massively parallel sequencing of footprints that actively translating ribosomes protect from nuclease digestion on messenger RNAs [15–18] , presents exciting opportunities to close this gap . Ideally , the millions of sequencing reads produced by a ribosome profiling experiment are snapshots of translation representing samples drawn from the steady state distribution of ribosomes across all coding sequences . The statistical properties of these snapshots can in theory be used to measure the relative speed with which each codon position is translated: the more often ribosomes are observed at a position , the longer ribosomes are inferred to spend at that position . In practice , ribosome profiling studies in Saccharomyces cerevisiae using different experimental protocols have reached contradictory conclusions about the average decoding times of codon identities . Because yeast rapidly regulate translation when stressed and ribosomes cannot be instantaneously harvested from cells , the original ribosome profiling protocol of Ingolia et al . [15] pretreats cells with cycloheximide ( CHX ) for several minutes to stabilize ribosomes in place before the harvesting process begins . CHX is a small-molecule translation inhibitor that has been a staple of experimental approaches to the study of translation for decades . The exact mechanism of this inhibition , however , is not completely understood , with recent studies suggesting that CHX binds to a ribosome’s E-site along with a deacylated tRNA to block further translocation [19 , 20] . The majority of the rapidly growing body of ribosome profiling experiments in yeast have followed this original CHX-pretreatment protocol [13 , 21–30] . Several groups have applied a variety of conceptually similar computational methods to the data produced by these experiments to infer the average speed with which each codon identity is translated . Counterintuitively , these groups have found that , on the whole , codons decoded by rare tRNAs appear to be translated faster than those decoded by more abundant tRNAs [31 , 32] . Different theories have been advanced to explain these unexpected results , hypothesizing that the measured elongation rates reflect a co-evolved balance between codon usage and tRNA abundances [31] , or that translation dynamics are dominated by interactions involving the nascent chain rather than the actual decoding process [32] . An alternative experimental protocol uses an optimized harvesting and flash-freezing process that allows CHX pretreatment to be omitted [15 , 17 , 26 , 29 , 33–38] . Experiments performed with this protocol have revealed that treatment with CHX affects several high-level characteristics of footprinting data , including the distribution of lengths of nuclease-protected fragments in mammalian cells [39] and the amount of enrichment in ribosome density at the 5’ end of coding sequences in yeast [34] . In contrast to data produced using CHX pretreatment , several studies using this alternative protocol have reported that non-optimal codons are in fact translated more slowly [33 , 38] . The source of this discrepancy between the statistical properties of measured ribosome positions with and without CHX pretreatment has been unclear , leading to uncertainty as to which measurements correspond to actual properties of in vivo translation dynamics . Here , we present analysis of data from a large body of ribosome profiling studies in yeast to resolve these contradictory results . We find consistent differences between experiments performed with and without CHX pretreatment in how often ribosomes are measured with specific codon identities positioned in their tRNA binding sites . We also find unexpected patterns in how often ribosomes are found downstream of specific codon identities in experiments using CHX pretreatment . Together , these observations suggest that translation elongation continues for many cycles after the introduction of CHX , but that the amount of time ribosomes spend translating each codon under these perturbed conditions is quite different from the unperturbed dynamics .
To characterize differences in the measured positions of ribosomes when cells are pretreated with CHX and when they are not , we compared the relative amount of time ribosomes spend at each codon position in data from many different ribosome profiling experiments . For each experiment , we mapped footprint sequencing reads to yeast coding sequences and assigned each read to the codon position in the A-site of the associated ribosome ( the sixth codon from the 5’ end in a canonical 28nt footprint [15] ) . In each coding sequence , we divided the raw count of reads at each codon by the average across the coding sequence to produce a relative enrichment value for each position . To measure the average relative amount of time a codon identity spends in the A-site of a ribosome each time it is translated , we computed the mean of these relative enrichment values at all occurrences of each codon across all yeast coding sequences ( Fig 1A ) . If the mean relative enrichment of a codon is higher than 1 , translation of the codon is inferred to be slower than the average speed of its surroundings . Conversely , a mean relative enrichment value lower than 1 indicates that translation of a codon is faster than its surroundings . We compared the mean relative A-site enrichments for all 61 non-stop codons between the original CHX-pretreatment data of Ingolia et al . [15] , an experiment we performed using the same CHX-pretreatment protocol , and data from experiments without CHX pretreatment by Gerashchenko [34] and by Weinberg et al . [38] ( Fig 1B ) . A-site occupancies are strongly positively correlated between experiments that use CHX pretreatment ( upper left panel ) and between experiments that do not ( lower right panel ) . The two sets of values reproducibly reported by each experimental protocol are inconsistent with each other , however , with a moderate negative correlation between them ( upper right panel ) . To test the generality of these comparisons , we computed Pearson correlations between the A-site occupancies in representative experiments from many different studies in yeast and performed unsupervised hierarchical clustering on the resulting matrix of correlation values ( Fig 1C ) . Experiments with and without CHX pretreatment separate into two distinct clusters , confirming that the two experimental conditions produce two reproducible but different pictures of translation dynamics . Because the codon located in the A-site is not expected to be the only determinant of how long a ribosome spends at each position , we also calculated the influence of the codon located in the P- or E-site of a ribosome on measured ribosome density in each experiment . To do this , for each codon identity , we computed the average relative enrichment of ribosomes whose A-site was one position ( P-site , S1A Fig ) or two positions ( E-site , S2A Fig ) downstream of the codon identity , rather than directly over the codon identity . Clustering the same set of experiments by the correlations between their P-site occupancy values for all 61 codons recapitulates the same strong separation of CHX experiments from no-CHX experiments produced by the A-site occupancies ( S1B and S1C Fig ) . E-site occupancy values exhibit less dynamic range from codon to codon under either experimental condition compared to the A- or P-sites ( S2B Fig ) but still generally separate the two conditions from each other ( S2C Fig ) . The fact that tRNA binding site enrichment values from experiments with and without CHX pretreatment separate into two clusters represents two incompatible claims about how long ribosomes spend translating each codon identity . To test how well each of these two apparent phenotypes agreed with intuitive expectations about elongation times , for each experiment , we computed the Spearman rank correlation between each codon identity’s mean relative A-site enrichment and the inverse of its tRNA adaptation index ( tAI ) [5 , 40] . The tAI of each codon identity is the weighted sum of the genomic copy numbers of the different tRNA genes that can decode the codon , with empirically determined weights penalizing wobble base pairings . This calculation quantifies the expectation that tRNAs expressed at lower abundances or that involve non-standard base pairing in their codon-anticodon interaction should require longer to translate . Consistent with previous reports [13 , 23 , 31 , 32] , all CHX experiments report weak to moderate negative correlations ( Fig 1D , orange labels ) , representing apparent translation dynamics in which less abundant tRNAs are actually translated faster . Experiments without CHX , on the other hand , report positive correlations of varying magnitude ( Fig 1D , 0x Gerashchenko NAR points and purple labels ) . Experiments by Pop [36] , Lareau [26] , Nedialkova [29] , Guydosh [35] and Gardin [33] produce weak to moderate correlations , but experiments by Gerashchenko [34] , Jan [41] , Williams [42] , Weinberg [38] , and Young [37] produce fairly strong and highly statistically significant correlations . Serendipitously , a series of experiments by Gerashchenko [34] performed to measure the effect of CHX concentration on the observed ramp in ribosome density at the 5’ end of coding sequences provide a way to confirm that CHX is directly responsible for these contradictory results . Gerashchenko produced datasets using pretreatment with a gradient of seven different CHX concentrations ( 0x , 1/64x , 1/16x , 1/4x , 1x , 8x , and 100x , expressed in multiples of the original protocol’s concentration of 100 μg/ml ) for two different cellular conditions ( unstressed and oxidatively stressed cells ) , and using two different concentrations ( 0x and 1x ) for heat shocked cells . Intriguingly , the rank correlation of A-site enrichment with 1 / tAI in these experiments moves smoothly from moderately negative with the highest CHX concentration to strongly positive with no CHX across each set of samples ( Fig 1D ) , with only one sample ( 1/16x unstressed ) deviating from perfect monotonicity . This concentration-dependence is a strong confirmation that CHX systematically shifts statistical properties of where ribosomes are measured . To further explore the effect of CHX on A-site occupancies , we plotted the movement of the mean relative A-site enrichment of each codon identity across the concentration gradient ( Fig 2 shows data from the oxidatively stressed set of samples; S3 Fig shows data from all three sets ) . Strikingly , a set of the codons with the highest enrichments ( that is , the codons that are slowest to translate ) when there is no CHX undergo consistent , gradual depletion with increasing concentration until they become among the fastest . Two prominent examples are CGA and CGG , codons encoding arginine . Mean relative enrichment at CGA codons is approximately four with no CHX , but this steadily decreases to a final value of less than 1/2 at the highest CHX concentration . CGA is translated by a tRNA identity with a moderate genomic copy number . However , its first anti-codon nucleotide is postranscriptionally modified to an inosine , making it the only codon in yeast that is decoded exclusively by an I-A wobble pairing [7] . Several studies have demonstrated that this leads to substantial translational pausing at occurrences of CGAs , particularly at CGA-CGA dicodons [10 , 12 , 43] . CGG , which is decoded by a tRNA with only a single genomic copy and therefore also expected to be slowly translated , undergoes an even larger shift from apparently slow with no CHX to apparently fast with high CHX concentration . For both codons , a CHX-concentration-dependent ( and therefore almost certainly CHX-mediated ) mechanism drives measured translation speeds away from their intuitively expected values . We also examined changes in occupancy at the P- and E-sites across the concentration gradient ( S3 Fig ) . A smaller number of codons undergo substantial changes in mean relative P-site enrichment , with the dominant effect being a dramatic reduction in CGA enrichment with increasing CHX concentration . Compared to the A- and P-sites , there is less concentration-linked change in occupancy at the E-site . Although there is no a priori reason to expect a codon to have any impact on the translation speed of a ribosome whose A-site is more than a few positions away from it , it is straightforward to measure the average ribosome density at any particular offset upstream or downstream of a given codon identity . To do this , relative enrichments of footprint reads with their A-site at each codon position are computed as above . The relative enrichments values at all codon positions located exactly the offset of interest away from an occurrence of the codon identity of interest are then averaged ( Fig 3A ) . We computed mean enrichment values for a wide range of offsets around each codon identity in data from our ( CHX pretreatment ) experiment . The A- , P- , and E-site occupancies above are the special cases of offsets of 0 , +1 and +2 downstream , respectively . Although we did not expect mean enrichments to deviate substantially from one at offsets that are far removed from the tRNA binding sites , we were surprised to find prominent peaks and dips in enrichments downstream of many codon identities ( Figs 3B and S4 ) . After observing these peaks in our data , we examined data from many other experiments in yeast for evidence of similar peaks . Peaks are ubiquitous in data from experiments using CHX pretreatment ( Figs 3C , 3D and S5 ) , but are almost entirely absent in data from experiments that do not use CHX pretreatment ( Figs 3C , 3D and S6 , but see discussion of Pop et al . data below ) . For data from a particular experiment , the peaks corresponding to different codon identities occupy roughly the same range of offsets downstream ( Figs 3B and S4 ) , but across experiments carried out by different groups , the locations and shapes of the set of peaks change considerably ( Figs 3C and S5 ) . The centers of peaks vary from as close as ∼7 codons downstream in data from Nedialkova [29] to as far away as ∼50 codons downstream in data from McManus [27] , with other CHX experiments densely populating the range of offsets between these observed extremes . Peaks become broader in width and smaller in maximum magnitude the farther downstream they are located . To test if CHX treatment had a concentration-dependent effect on the locations and shapes of these peaks , we again turned to data from the CHX concentration gradient experiments of Gerashchenko . Fig 3D shows enrichment profiles downstream of CGA in one series of samples; S7 Fig shows profiles in all three series . Peaks are absent in the samples with no CHX and minimal in the samples with concentrations below 1/4x the standard concentration ( with the notable exception of unstressed 1/16x , which is also a clear outlier in Figs 1D and S3 ) . For samples with concentrations greater than or equal to 1x , for which clear peaks are observed , peaks are located less far downstream and become narrower and taller as CHX concentration increases ( Fig 3D ) . Other studies have hypothesized that interactions between recently incorporated amino acids and the ribosome exit tunnel lead to slower ribosome movement downstream of occurrences of certain amino acids [32 , 44] . There are two lines of evidence that the downstream peaks observed here are not simply the result of these effects . First , within a single sample , the magnitudes of peaks vary substantially between different codons encoding the same amino acid . As examples , the peak downstream of CGA in our experiment is substantially higher than the peaks downstream of other codons encoding arginine ( Figs 3B and S4 ) , and GCG is the only codon encoding alanine for which there is an appreciable downstream peak ( S4 Fig ) . If these peaks were caused by interactions involving an amino acid in the nascent polypeptide chain , they should be agnostic to the codon identity used to encode the amino acid . Second , the facts that the locations of peaks change in response to changes in CHX concentration and that peaks disappear in the absence of CHX strongly suggest that the peaks are a consequence of CHX pretreatment rather than a genuine feature of translation . Having observed large shifts in tRNA binding site occupancies between experiments with and without CHX pretreatment and the appearance of downstream peaks in CHX experiments , we sought a model for how CHX treatment disrupts the measured positions of ribosomes that could parsimoniously explain both phenomena . To test potential models , we developed a simulation of the movement of ribosomes along yeast coding sequences; see supplement for simulation details . By incorporating different possible effects of the introduction of CHX into these simulations , we could evaluate the ability of different models to explain the observed features of the experimental data . A natural first hypothesis is that each ribosome waits an exponentially distributed amount of time until a CHX molecule diffuses into the ribosome’s E-site and irreversibly arrests it , with shorter average waiting times when increased CHX concentration is used . If ribosomes continue to spend the same relative amounts of time on each codon while waiting for CHX to arrive , however , the position of each ribosome at the random instant of CHX arrival samples from the same steady state distribution that ribosomes occupied before CHX was introduced . We confirmed by simulation that this potential mechanism produces neither downstream peaks nor substantial changes in A-site occupancies . We therefore considered the unexpected hypothesis that ribosomes continue to advance from codon to codon after beginning to interact with CHX , but that the relative elongation rate of each codon during this continued elongation is substantially changed from its value before CHX treatment . To model this possible behavior , we simulated translation with the relative elongation time of each codon set to its mean relative A-site enrichment as measured in a no-CHX-pretreatment experiment until steady state was reached . We then switched the relative elongation time of each codon to its A-site enrichment as measured in a CHX-pretreatment experiment and allowed translation to proceed under these new elongation rates for a short period of time . At the end of this short period of continued elongation , we recorded the positions of all ribosomes and processed the simulated ribosome footprints identically to the real experimental datasets . Interestingly , the resulting simulated enrichment profiles around different codons qualitatively reproduce both major phenomena observed in data from CHX experiments ( Fig 4A ) . To better understand why changing relative elongation rates shortly before measuring ribosome positions produces these patterns , we constructed a simple analytical model of the translation of many copies of a particular coding sequence . In this model , ribosomes wait an exponentially distributed amount of time at each codon position before moving on to the next , with the mean of this exponential distribution depending only on the codon in the A-site of the ribosome . This implies that the steady state ribosome density at each position is proportional to the mean elongation time of the position; see supplement for details . We considered a hypothetical coding sequence consisting of one codon that is translated slowly surrounded on either side by many identical copies of a codon that is translated faster . We computed the density of ribosomes across this coding sequence at steady state under these elongation dynamics , producing the expected excess in density at the slow codon ( Fig 4B , red ) . Then , at t = 0 ( arbitrary units ) , we changed the relative elongation rates of the two codon identities so that the previously slow codon was now faster than its surroundings . We analyzed the evolution of ribosome density across the coding sequence over time following this change . After a sufficiently long time , the system will have reached the steady state of the new dynamics , in which ribosome density is lower at the now-faster codon than its uniform level at all of the surrounding codons . Immediately after the rates are changed , however , ribosomes are still distributed at the steady state densities implied by the old dynamics and are therefore out of equilibrium under the new dynamics . There is a temporary excess of ribosomes at the formerly-slow codon , and the process of relaxing from the old steady state to the new steady state manifests as these excess ribosomes advancing along the coding sequence over time ( Fig 4B ) . Stochastic variation in the wait times of each individual ribosome at each subsequent codon position causes the excess to gradually spread out as it advances . Hypothetical measurements of the positions of all ribosomes at a series of increasing times after the change to the new dynamics would therefore produces patterns that look like an advancing wave of enrichment , as is seen around e . g . several arginine codons in real ( Fig 3 ) and simulated ( Fig 4A ) data . We also considered a hypothetical coding sequence in which a single special codon undergoes an increase , rather than decrease , in relative elongation time compared to stretches of identical codons on either side ( S8 Fig ) . In this case , the time period immediately following the change in dynamics is spent filling the formerly-faster codon position up to its newly increased steady state density . During this time , there are temporarily fewer ribosomes being promoted onward to downstream positions than there were before the change . This results in a transient wave of depletion , rather than enrichment , that advances away from the formerly-faster codon position and spreads out over time ( S8B Fig ) . This is qualitatively consistent with the profile of depletions downstream of e . g . two isoleucine codons in real ( S4 Fig ) and simulated ( S8A Fig ) data . The hypothesis that changes in measured tRNA binding site occupancies and the appearance of downstream peaks are both caused by continued elongation with disrupted dynamics in the presence of CHX makes a testable prediction about the quantitative link between these two phenomena . If downstream peaks are transient waves moving downstream after a change in the relative amounts of time ribosomes spend positioned over each codon , the total CHX-induced excess or deficit in enrichment downstream of each codon identity should exactly offset the total CHX-induced change in enrichments at the tRNA binding sites . To test whether experimental data agreed with this prediction , we analyzed several matched pair of experiments performed with and without CHX by Jan [41] ( Fig 5 ) , Williams [42] , and Gerashchenko [34] ( S10 Fig ) . For each codon identity , we compared the sum of the differences in enrichment between the experiments at the A- , P- , and E-sites ( green area in insets ) to the sum of the difference in enrichment across the range of downstream offsets occupied by the putative waves ( red area in insets ) . In all matched pairs of experiments , the area of each codon’s downstream peak is accurately predicted by its tRNA binding site changes ( r2 = 0 . 85 to 0 . 93 , slope of best fit line β = −1 . 00 to −1 . 20 ) . Insets in Figs 5 and S9 plot the changes in tRNA binding site enrichments and downstream wave areas for four different codons to demonstrate the full range of agreement between the two phenomena . CGA , which undergoes comparably large decreases in enrichment at both the A- and P-sites , produces a downstream wave with approximately twice the area of CCG , which undergoes a large decrease in enrichment at the A-site but not the P-site . Codons with similar enrichments at all three tRNA binding sites between the two experiments , such as ACT , produce no appreciable downstream waves , while several codons that undergo modest increases in enrichment at the binding sites , such as TTG , produce proportionally modest net deficits of enrichment downstream . This close correspondence strongly suggests that the downstream peaks are in fact transient waves , and therefore that tRNA binding site enrichments in CHX experiments do not reflect natural translation dynamics . A model of continuing elongation after CHX has begun interacting with ribosomes also predicts that CHX pretreatment for increasing amounts of time under otherwise identical conditions should produce waves that have advanced proportionally farther downstream . To test this prediction , we examined mean relative enrichments around CGA in the full set of experiments from Jan et al . [41] , which involved a variety of different combinations of CHX pretreatment duration and temperature . Strikingly , when pretreatment was performed at 30°C , downstream peaks after 9 minutes of pretreatment are centered slightly more than twice as far downstream as those after 4 minutes of pretreatment ( Fig 6 ) , further supporting the hypothesis that elongation continues to occur at a slow but steady rate over the course of CHX pretreatment . Under this model of continuing elongation , the areas of downstream waves can be used to recover indirect information about the history of translation dynamics in each CHX-pretreatment experiment before these dynamics were disrupted by CHX . Specifically , we can estimate what the sum of the enrichments at the A- , P- , and E-sites was for each codon before the introduction of CHX by adding the net area of the wave that moved downstream during elongation in the presence of CHX back to the sum of the enrichments that remain at the binding sites . We will call this quantity the corrected aggregate enrichment of each codon . It can be interpreted as the average relative amount of time that a ribosome took to decode each occurrence of a codon before CHX was introduced , from when the codon was presented in the A-site to when it left the E-site . While we would prefer to recover how long each codon spent in each individual tRNA binding site in these experiments , this single-codon-resolution information has been irreversibly lost . As the CGA and CCG insets in Fig 5 demonstrate , changes in enrichment at the A-site or at the P-site result in downstream waves that occupy the same large range of downstream offsets , so the area in each wave cannot be unambiguously assigned back to a particular tRNA binding site . To test if codons decoded by less abundant or wobble-paired tRNAs tended to be translated more slowly than more abundant tRNAs in CHX experiments before the introduction of CHX , we computed the Spearman rank correlation between the corrected aggregate enrichment of each codon identity and 1 / tAI . Corrected aggregate enrichment correlates positively with 1 / tAI for every CHX experiment analyzed ( Fig 7 , purple dots ) , recovering an intuitively expected signature of translation dynamics that is absent in CHX experiments if the total elongation time is estimated by the sum of the tRNA binding sites enrichments alone ( Fig 7 , green dots ) . Continued elongation with disrupted dynamics after the introduction of CHX also offers a potential explanation for counterintuitive results in a set of experiments by Zinshteyn et al . [23] . Zinshteyn performed ribosome profiling on yeast strains that lacked different genes required to post-transcriptionally add mcm5s2 groups to a uridine in the anticodons of tRNAs that decode codons ending in AA and AG . These anticodon modifications are thought to enhance codon-anticodon recognition and speed up translation of these codons [45] . Surprisingly , Zinshteyn found that measured changes in tRNA binding site occupancies between deletion strains and the wild type were much smaller than expected given the phenotypic consequences of lacking these modifications . These experiments followed the standard CHX pretreatment protocol , however , and we observe clear downstream waves in enrichment in all of them ( S11 Fig ) . According to our model , therefore , tRNA binding site occupancy levels in these experiments reflect properties of elongation in the presence of CHX rather than of in vivo dynamics . To test if CHX-disrupted elongation was masking the true impact of the absence of anticodon modification in these experiments , we compared the profiles of mean enrichment around all codon identities decoded by the modification-deficit tRNA species between the deletion strains and the wild type . Intriguingly , the profiles of mean enrichment around AAA showed consistently increased downstream wave areas in all of the deletion strains compared to wild type ( Fig 8A ) . To quantify this increase , we computed the corrected aggregate enrichment of each codon identity as above by adding the downstream wave area to the sum of the binding site enrichments . We then computed the change in corrected aggregate enrichment for each codon identity between each of the deletion strains and the wild type ( Figs 8B and S12 ) . In each of the deletion strains , but not in a replicate of the wild type , AAA undergoes a dramatically larger increase in aggregate tRNA binding site enrichment when corrected to include downstream wave area ( purple ) than if wave area is not included ( green ) . This argues that AAA does in fact take substantially longer to decode in vivo in cells lacking the ability to modify its tRNA , but that most of this difference disappears during continued elongation in the presence of CHX .
Although the exact mechanistic details of how disrupted elongation in the presence of CHX occurs remains unclear , there are several key features of observed patterns in the data and of known properties of CHX that any potential mechanism must accommodate . The first is that the disruption in dynamics is concentration-dependent . The second is that relative elongation rates in the new dynamics are still coupled to codon identities . Mean relative A- and P-site enrichments do not simply collapse towards being uniform in CHX experiments , but instead reproducibly take on a wide dynamic range of codon-specific values . The third is that absolute elongation rates must be dramatically slower in the presence of standard concentrations of CHX . Any model implying the contrary is not plausible; CHX has been successfully used as a translation inhibitor for decades . We suggest that this inhibition is accomplished by a large reduction in the rate of elongation rather than a complete halt . A possible mechanism with all three of these properties is that CHX repeatedly binds and unbinds ribosomes , preventing advancement when bound but allowing elongation to proceed when unbound . In this model , the global rate of CHX binding to all ribosomes increases with increasing CHX concentration , leading to a decrease in the amount of time each ribosome spends unbound and therefore globally decreasing the rate of continued elongation . This accounts for the fact that downstream peaks move less far downstream in the same amount of time with increasing CHX concentration ( Fig 3D ) . For fixed CHX concentration , there is close proportionality between the duration of CHX pretreatment and the distance peaks have moved downstream in experiments performed by the same group ( Fig 6 ) . While comparisons of experiments by different groups may be confounded by subtle differences in experimental protocols , there is broad ( but not perfect ) agreement between downstream peak distance and annotated pretreatment time across experiments from different studies , with longer pretreatment times ( e . g . 5 minutes in McManus [27] ) corresponding to peaks farther downstream and shorter pretreatment times ( e . g . 1 minute in Lareau [26] and Nedialkova [29] ) corresponding to closer peaks . Because the distance that peaks have moved downstream is the product of the total duration of disrupted elongation and the average rate of this elongation , the magnitude of the reduction in elongation rate can be roughly estimated . The range of peak centers with standard CHX concentration and 2 minutes of pretreatment implies absolute elongation rates during CHX treatment of 0 . 1 to 0 . 3 aa/s . Because natural elongation rates in yeast are 7 to 9 aa/s [46] , this represents an approximately 20- to 90-fold reduction in the speed of elongation . To explain the reproducible range of codon-specific elongation rates in the presence of CHX , changes in the conformation of the ribosome as a result of differences in the geometry or base-pairing interactions of the tRNAs occupying the A- and P-sites could modulate the rates of CHX binding and unbinding . Conformational changes in the ribosome as a result of codon-anticodon interactions are known to be an integral part of the elongation cycle [1] . Given the unique presence of I-A wobble pairing in the decoding of CGA codons , the outsize role that CGA plays in these phenomena suggests that base-pairing interactions could play a major role in determining CHX affinity . This offers an elegant potential explanation for the negative correlation between A-site enrichments with and without CHX: codon identities that produce unusual ribosome conformations tend to slow down elongation when tRNA binding is rate-limiting , but tend to speed up elongation when CHX disassociation is rate-limiting . In this model , the concentration-dependent interpolation between these two regimes observed in Fig 2 reflects the fact that as CHX concentration decreases , each ribosome spends an increasing fraction of time unbound by CHX and therefore elongating according to the unperturbed dynamics . Heterogeneity in measured A- and P-site enrichment values between experiments that avoid CHX pretreatment presents complications for this model . Except for two experiments in the set performed by Guydosh [35] , however , all such experiments still include CHX in the lysis buffer into which cells are harvested . Under some conditions , the same continued elongation with disrupted dynamics that occurs during CHX pretreatment could also occur during the harvesting process once ribosomes have been exposed to CHX in the lysis buffer . The A- and P-site occupancies in the non-pretreated experiments by Pop , Lareau , Gardin , Nedialkova , and the unstressed experiment by Gerashchenko can be interpreted as an intermediate phenotype halfway in between the two tighter clusters consisting of CHX-pretreatment experiments and of the non-pretreated experiments by Weinberg , Jan , Williams , Guydosh , Young , and the other two experiments by Gerashchenko ( Fig 1B ) , potentially reflecting a small amount of CHX-mediated elongation in these intermediate experiments . Consistent with this interpretation , enrichment profiles around CGA and CGG appear to be shifted slightly downstream in these intermediate-phenotype non-pretreated experiments ( S13 and S15 Figs ) . The complete set of five non-pretreated experiments produced by Pop et al . [36] are particularly heterogeneous . Three of these experiments ( WT-URA_footprint , AGG-OE_footprint , and AGG-QC_footprint ) report A-site occupancies strikingly similar to values reported by CHX pretreatment experiments and less similar to the other non-pretreated experiments ( S14 Fig ) . These same three experiments also have distinct peaks located approximately 20 codon positions downstream in the enrichment profiles around each codon identity that are absent in the other two experiments from the study ( S15 Fig ) . Such extreme heterogeneity between non-pretreated experiments is difficult to account for in our model , but suggests that a wide range of different amounts of elongation after exposure to the lysis buffer are possible across different implementations of harvesting protocols . In light of our model , several counterintuitive results from previous ribosome profiling studies in yeast take on new interpretations . Most notably , we offer an explanation for contradictory claims about whether so-called optimal codons corresponding to more abundant tRNAs are translated more rapidly . In many organisms , optimal codons are used with greater frequency in highly expressed genes , and a large body of theoretical work assumes that increased elongation speed drives this tendency [14] . If optimal codons are decoded faster , this tendency could lead directly to increased expression by increasing the rate of production of protein from each message or by avoiding mRNA-decay pathways linked to ribosome pausing [47] . Alternatively , if translation initiation is typically rate-limiting , using faster codons reduces the amount of time ribosomes spend sequestered on highly transcribed mRNAs , freeing up ribosomes to translate other messages and leading to more efficient system-wide translation [48] . If A-site enrichments measured in CHX pretreatment experiments reflected natural translation dynamics , however , optimal codons would not be elongated more quickly , and these theories fall apart . By offering a model for why the measured positions of ribosomes in CHX experiments appear to report that non-optimal codons are the fastest to be translated , and by showing evidence that optimal codons were in fact being translated more quickly before the introduction of CHX , we enable a principled resolution to this controversy . Earlier studies in this area have hypothesized that the A-site enrichments reported by CHX experiments could reflect an optimal balance between codon usage and tRNA abundance [31] , or that potential heterogeneity in elongation times at different occurrences of the same codon identity could conspire to produce these A-site enrichments [49] . By showing that the A-site occupancies fed into these models almost certainly do not represent the actual in vivo dynamics , our results argue against both of these conclusions . Finally , our model offers an explanation for the small apparent impacts of several experimental attempts to modify tRNA repertoires on measured relative elongation rates . CHX-pretreatment experiments by Zinshteyn et al . [23] on mcm5s2U-pathway deletion strains show surprisingly small changes in tRNA binding site occupancies at codons decoded by the modification-deficient tRNAs . Our observation that all of the deletion strains have substantially increased waves of enrichment downstream of AAA , one such codon identity , compared to wild type suggests that binding site occupancy changes in the presence of CHX dramatically underestimate the actual in vivo increase in the decoding time of AAA in the deletion strains . Pop et al . [36] evaluated the impact of overexpressing , deleting , or modifying the body sequence of tRNAs and also found surprisingly small changes in the rates at which the corresponding codon identities were translated . As discussed above , the experiments of Pop et al . did not pretreat with CHX , but a subset of these experiments show both clear downstream peaks and A-site occupancies shifted towards values reported by CHX pretreatment experiments . This suggests that enough CHX-disrupted elongation occurred in these experiments during the harvesting process that the resulting A-site occupancies may not be able to measure any potential effects of the tRNA repertoire modifications . Repeating these experiments without any CHX in order to accurately sample from in vivo dynamics could clarify the consequences of these tRNA manipulations .
Saccharomyces cerevisiae strain AJY3373 ( MATα , KanMX-PGAL1-RPL10 , his3Δ1 , ura3Δ0 , leu2Δ0 ) with plasmid pAJ2522 ( WT RPL10 , LEU2 CEN ) was grown in 250ml YPD to mid-log phase , shaking at 30°C . Before harvesting , cultures were treated with cycloheximide ( 100 μg/ml ) for an additional 2 min , shaking at 30°C . Cells were harvested by vacuum filtration in 0 . 45μm PES bottle top filters . Filtration took approximately 1 min per 250 ml culture , and cells were immediately scraped from the filter and plunged into liquid nitrogen . Frozen cell pellets were lysed by cryogenic grinding in a mixer mill and thawed in the presence of 1ml lysis buffer ( 20mM Tris pH7 . 4 , 150mM NaCl , 5mM MgCl2 , 1mM DTT , 100μg/ml cycloheximide , 1% Triton-X ) . Lysates were clarified by centrifugation and 25 A260 units of extract were treated with 7 . 5μL RNase I ( Ambion ) for 1 hour at room temperature . Ribosomes were isolated with Sephacryl S-400 micro spin columns , pre-equilibrated with buffer ( 20mM Tris pH7 . 4 , 150mM NaCl , 5mM MgCl2 ) . RNA was recovered from the eluate using an miRNeasy kit ( Qiagen ) . Deep sequencing libraries were generated from 27–29 nt footprint fragments following the protocol in [17] . We followed the rRNA depletion method from the same protocol , but used two yeast-specific rRNA oligos: 5’-ATTGCTCGAATATATTAGCATGGAATAATAGAATA-3’ 5’-AAGAGGTGCACAATCGACCGATCCTG-3’ Libraries were sequenced on an Illumina HiSeq . Briefly , for each experiment , adapter sequences were trimmed and reads mapping to yeast rRNA sequences were filtered out with bowtie2 . Remaining reads were aligned to the yeast genome and spliced transcript models with tophat2 . In each coding sequence , uniquely mapped reads were assigned to the codon in the A-site of the ribosome following [15] . Reads of length 28 or 29 were assigned to the in-frame codon closest to the nucleotide at ( 0-based ) offset +15 from the 5′ end of the read; reads of length 30 were assigned to the in-from codon closest to offset +16 . See supporting text for full details . For each experiment , after assigning read counts to A-site positions as described above , read counts were normalized within each coding sequence by dividing all counts by the mean across the coding sequence to produce relative enrichments . For each codon identity , for a range of offsets from -90 to 90 codon positions , the set of all codon positions across all coding sequences was then stratified to select those positions that are located at the offset of interest from an occurrence of the codon identity of interest , and the mean of the relative enrichment values at all positions in this stratified set is computed . To remove the influence of poorly-understood structure in measured ribosome density at the 5′ end of coding sequence , we excluded 90 codons at the beginning and end of each coding sequence from all mean relative enrichment computations . For all calculations involving downstream wave areas , we increased the number of codons excluded from the edges of genes from 90 to 200 . Excluding this wider range marginally improves the agreement between CHX-induced tRNA binding site changes and downstream wave areas . This suggests that patterns in codon composition , which are particularly pronounced at the beginning of genes [5] , may introduce small confounding biases in ribosome density around different codon identities that aggregate when adding relative enrichments over a wide range of offsets to calculate downstream wave areas . Clusterings of matrices of Pearson correlations between A- , P- , and E-site enrichments in Figs 1 , S1 and S2 were computed via UPGMA using Euclidean distances . | Ribosome profiling measures the precise locations of millions of actively translating ribosomes on mRNAs . In theory , the frequency with which ribosomes are observed positioned over each type of codon can be used to quantify the speed with which each codon is translated . In practice , ribosome profiling experiments in yeast that use translation inhibitors to arrest translation before measuring the positions of ribosomes report very different apparent translation speeds for each codon than experiments that do not use inhibitors . To explain this inconsistency , we show that a previously unappreciated mechanism causes experiments using translation inhibitors to not measure ribosomes at each position on mRNAs in proportion to the actual amount of time spent there in vivo . Understanding this mechanism reveals that experiments without inhibitors more accurately measure translation dynamics and provides guidance for the design and interpretation of future ribosome profiling experiments . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Understanding Biases in Ribosome Profiling Experiments Reveals Signatures of Translation Dynamics in Yeast |
Although more than 20 genetic susceptibility loci have been reported for type 2 diabetes ( T2D ) , most reported variants have small to moderate effects and account for only a small proportion of the heritability of T2D , suggesting that the majority of inter-person genetic variation in this disease remains to be determined . We conducted a multistage , genome-wide association study ( GWAS ) within the Asian Consortium of Diabetes to search for T2D susceptibility markers . From 590 , 887 SNPs genotyped in 1 , 019 T2D cases and 1 , 710 controls selected from Chinese women in Shanghai , we selected the top 2 , 100 SNPs that were not in linkage disequilibrium ( r2<0 . 2 ) with known T2D loci for in silico replication in three T2D GWAS conducted among European Americans , Koreans , and Singapore Chinese . The 5 most promising SNPs were genotyped in an independent set of 1 , 645 cases and 1 , 649 controls from Shanghai , and 4 of them were further genotyped in 1 , 487 cases and 3 , 316 controls from 2 additional Chinese studies . Consistent associations across all studies were found for rs1359790 ( 13q31 . 1 ) , rs10906115 ( 10p13 ) , and rs1436955 ( 15q22 . 2 ) with P-values ( per allele OR , 95%CI ) of 6 . 49×10−9 ( 1 . 15 , 1 . 10–1 . 20 ) , 1 . 45×10−8 ( 1 . 13 , 1 . 08–1 . 18 ) , and 7 . 14×10−7 ( 1 . 13 , 1 . 08–1 . 19 ) , respectively , in combined analyses of 9 , 794 cases and 14 , 615 controls . Our study provides strong evidence for a novel T2D susceptibility locus at 13q31 . 1 and the presence of new independent risk variants near regions ( 10p13 and 15q22 . 2 ) reported by previous GWAS .
Type 2 diabetes ( T2D ) is a common complex disease that affects over a billion people worldwide [1] . Through genome-wide association studies ( GWAS ) , at least 24 genetic susceptibility loci have been reported for T2D [1]–[9] , including a SNP , rs7593730 , at 2q24 near the RBMS1 and ITGB6 genes that was associated with diabetes risk in a recent report from the Nurses' Health Study/Health Professionals Follow-up Study ( NHS/HPFS ) [2] . However , most of the reported genetic variants have small to moderate effects and account for only a small proportion of the heritability of T2D , suggesting that the majority of inter-person genetic variation in this disease remains to be determined . Over the last two decades , China , like many other Asian countries , has experienced a dramatic increase in T2D incidence . Cumulative evidence suggests that Asians may be more susceptible to insulin resistance compared with populations of European ancestry [10] . However , among the previously reported T2D genetic markers , only three SNPs – including two reported very recently – have been identified in populations of Asian ancestry [8] , [9] . SNP rs2283228 in the KCNQ1 gene was identified in a 3-stage study that included 194 diabetes patients and 1 , 558 controls and 268 , 068 SNPs in the first ( discovery ) stage [8] . A study conducted among Han Chinese in Taiwan recently identified two additional novel loci in the protein tyrosine phosphatase receptor type D ( PTPRD; P = 8 . 54×10−10 ) and serine racemase ( SRR; P = 3 . 06×10−9 ) genes [9] . Large genetic studies conducted in Asian populations will facilitate the identification of additional genetic markers for T2D , particularly for markers with a higher frequency in Asians than in other populations . We recently completed a GWAS of T2D in Shanghai . We report here our first effort , using a fast-track , multiple-stage study approach , to identify novel genetic markers for diabetes .
The study protocol was approved by the institutional review boards at Vanderbilt University Medical Center and at each of the collaborating institutes . Informed consent was obtained from all participants . This study consisted of a discovery stage and two validation stages , i . e . an in silico and a de novo validation study . The overall study design is presented in Figure S1 . The discovery stage included 1 , 019 T2D cases , 886 incident T2D cases from the Shanghai Women's Health Study ( SWHS ) , an ongoing , population-based , prospective cohort study of women living in Shanghai , and 133 prevalent T2D cases identified among controls of the Shanghai Breast Cancer Study ( SBCS ) , who were recruited in Shanghai during approximately the same period as the SWHS [11] . Controls for the discovery phase were 1 , 710 non-diabetic female controls from the SBCS ( for further details , see Text S1 , online ) . The biologic samples used for genotyping in this study were collected by the SWHS and SBCS . DNA samples were genotyped using the Affymetrix Genome-Wide Human SNP Array 6 . 0 . Extensive quality control ( QC ) procedures were implemented in the study . In the SWHS/SBCS GWAS scan , three positive QC samples purchased from Coriell Cell Repositories and a negative QC sample were included in each of the 96-well plates of the Affymetrix SNP Array 6 . 0 . SNP data obtained from positive quality control samples showed a very high concordance rate of called genotypes based on 79 , 764 , 872 comparisons ( mean , 99 . 87%; median , 100% ) . Samples with genotyping call rates less than 95% were excluded . The sex of all study samples was confirmed to be female . The identity-by-descent analysis based on identity by state was performed to detect first-degree cryptic relationships using PLINK version 1 . 06 [12] . We excluded from the study 21 samples that had: 1 ) call rate <95% ( n = 5 ) ; 2 ) samples that were contaminated or had mixed-up labels or that had been duplicated ( n = 12 ) ; 3 ) first-degree relatives , such as parent-offspring or full siblings ( n = 4 ) . We also excluded from the analysis SNPs that met any of following criteria: 1 ) MAF <0 . 05; 2 ) call rate <95%; 3 ) P for Hardy-Weinburg equilibrium HWE <0 . 00001 in either the case or control groups or in the combined data set; 4 ) concordance rate <95% among the duplicated QC samples; 5 ) significant difference in allele frequency distribution ( P<0 . 00001 ) between the 886 T2D cases from the SWHS and the 133 T2D cases from the SBCS; 6 ) significant difference in missing rates between cases and controls ( P<0 . 00001 ) . After applying the QC filter , 590 , 887 SNPs remained for the analyses . Because of financial constraints , we conducted a fast-track validation study using an approach that combined in silico and de novo replication . We selected a total of 2 , 100 SNPs from the discovery phase that had P-values of 1 . 3×10−9 to 5 . 0×10−3 derived from the additive model and that were not in linkage disequilibrium ( LD; r2<0 . 2 based on the HapMap CHB dataset ) with any previously reported T2D GWAS SNPs for an in silico replication using the GWAS scan data from the NHS/HPFS [2] . We used the NHS/HPFS T2D GWAS scans for our first step of validation , because the Shanghai T2D GWAS was conducted concurrently and used the same genotyping platform as the NHS/HPFS T2D GWAS and a priori arrangement was made for the two studies to exchange the top 2 , 000 SNPs for in silico replication . The NHS/HPFS T2D GWAS included 2 , 591 cases and 3 , 052 controls of European ancestry . We recognize that this approach may have reduced our chances of finding ethnicity-specific T2D markers , however , this approach had the advantage of enhancing our ability of finding true genetic markers . From the first in silico replication , 65 SNPs with the same direction of association in both studies and with a MAF >20% were chosen for a second in silico replication using GWAS scan data from a Korean T2D study , which included 1 , 042 cases and 2 , 943 controls genotyped with the Affymetrix Genome-Wide Human SNP Array 5 . 0 platform . In order to improve yield , only the top SNPs that are included in Affymetrix 5 . 0 ( N = 56 ) or that are in high LD ( r2>0 . 8 ) with at least one SNP on Affymetrix 5 . 0 ( N = 9 ) were selected for replication ( Table S1 ) . Of the 65 SNPs , the top 8 SNPs replicated in the Korean T2D study were further investigated using GWAS data from a T2D study conducted among Singapore Chinese ( 2 , 010 cases and 1 , 945 controls ) who were genotyped by using Illumina HumanHap 610 or Illumina Human1M ( Table S2 ) . Four of the 8 SNPs were not directly genotyped in the Singapore study , so instead , we selected SNPs that are in strong LD with these 4 SNPs ( imputed SNP information became available recently and is presented in this report ) . Finally , the 5 top SNPs ( rs2815429 , rs10906115 , rs1359790 , rs10751301 , and rs1436955 ) were selected for de novo genotyping in an independent sample set of 1 , 645 T2D cases and 1 , 649 controls identified from the SWHS and Shanghai Men's Health Study ( SMHS ) . Four of these SNPs ( rs10906115 , rs1359790 , rs10751301 , and rs1436955 ) were selected for the final stage of de novo genotyping replication in two independent Chinese studies , the Wuhan Diabetes Study ( WDS; 1 , 063 cases and 1 , 408 controls ) and the Nutrition and Health of Aging Population in China ( NHAPC ) study ( 424 cases and 1 , 908 controls ) . Detailed descriptions of the study designs and populations for each of the participating studies are presented in Text S1 online . Genotyping for the 5 SNPs included in the SWHS and SMHS sample set was completed using the iPLEX Sequenom MassArray platform . Included in each 96-well plate as quality control samples were two negative controls , two blinded duplicates , and two samples included in the HapMap project . We also included 65 subjects who had been genotyped by the Affymetrix SNP Array 6 . 0 in the Sequenom genotyping . The consistency rate was 100% for all SNPs for the blinded duplicates , compared with the HapMap data and compared with data from the Affymetrix SNP Array 6 . 0 . Genotyping for the final 4 SNPs in the WDS and NHAPC was completed using TaqMan assays at the two local institute laboratories using reagents provided by the Vanderbilt Molecular Epidemiology Laboratory . Both laboratories were asked to genotype a trial plate provided by the Vanderbilt Molecular Epidemiology Laboratory that contained DNA from 70 Chinese samples before the main study genotyping was conducted . The consistency rates for these trial samples were 100% compared with genotypes previously determined at Vanderbilt for all four SNPs in both local laboratories . In addition , replicate samples comparing 3–7% of all study samples were dispersed among genotyping plates for both studies . The imputation of un-genotyped SNPs in all participating GWASs was carried out after the completion of the current study using the programs MACH ( http://www . sph . umich . edu/csg/abecasis/MACH/ ) or IMPUTE ( https://mathgen . stats . ox . ac . uk/impute ) with HapMap Asian data as the reference for Asians and CEU data as the reference for European-ancestry samples . Only data with high imputation quality ( RSQR >0 . 3 for MACH ) were included in the current analysis . PLINK version 1 . 06 was used to analyze genome-wide data obtained in the SBCS/SWHS GWAS scan . Population structure was evaluated by principal component analysis using EIGENSTRAT ( http://genepath . med . harvard . edu/~reich/Software . htm ) . A set of 12 , 533 SNPs with a MAF ≥10% in Chinese samples and a distance of ≥25 kb between two adjacent SNPs was selected to evaluate the population structure . The first two principal components were included in the logistic regression models for adjustment of population structures . The inflation factor λ was estimated to be 1 . 03 , suggesting that population substructure , if present , should not have any appreciable effect on the results . Pooled and meta-analyses were carried out in SAS to derive combined odds ratios ( OR ) by using data from studies of all stages . We applied the weighted z-statistics method , where weights are proportional to the square root of the number of subjects in each study . Results from both random and fixed effect models are presented . ORs and 95% confidence intervals ( CI ) were estimated using logistic regression models with adjustment for age , BMI , population structure ( for GWAS data ) , and gender , when appropriate . Analyses with additional adjustment for smoking were conducted by pooled analysis whenever possible and by meta-analysis when KARE data were included in order to examine the confounding and modification effects of these factors ( Table S2 ) . Genotype distributions for the top 4 SNPs included in the final de novo genotyping were consistent with HWE ( P> 0 . 05 ) in each study . All P values presented are based on two-tailed tests , except where indicated otherwise .
The general characteristics of the participating study populations are presented in Table 1 . T2D cases had a higher BMI than controls across all studies . Except for the SWHS , SMHS , and Shanghai Nutrition Institute ( SNI ) validation studies , where cases and controls were matched on age , cases were older than controls in all other studies . A difference in gender distribution was also seen in several studies . These variables were adjusted for in subsequent analyses . Table 2 presents the results of analyses of associations of T2D with previously reported , GWAS-identified genetic markers in our discovery samples [1]–[9] . Of the 24 SNPs reported by previous GWAS , 15 were directly genotyped by the Affymetrix SNP Array 6 . 0 . One SNP ( rs7578597 ) showed a MAF = 0 in HapMap CHB data and was not included on the Affymetrix 6 . 0 chip . The remaining 8 SNPs , including rs2943641 , rs10010131 , rs13266634 , rs12779790 , and rs4430796 , as well as the newly identified markers rs391300 and rs17584499 , were imputed . SNP rs4430796 showed low imputation quality ( RSQR = 0 . 06 ) in the SBCS/SWHS GWAS and was excluded from the analysis . We found that 8 of these SNPs showed an association consistent with initial reports at P<0 . 05 , including rs4402960 ( 3q27 . 2 , IGF2BP2 ) , rs10946398 ( 6p22 . 3 , CDKAL1 ) , rs13266634 ( 8q24 . 11 , SLC30A8 ) , rs10811661 ( 9p21 . 3 , CDKN2A/B ) , rs5015480 ( 10q23 . 33 , HHEX ) , rs7901695 ( 10q25 . 2 , TCF7L2 ) , rs2283228 ( 11p15 . 5 , KCNQ1 ) , and rs5215 ( 11p15 . 1 , KCNJ11 ) . Among the remaining 11 SNPs , 4 SNPs had a MAF of 3–7% in our study population . Thus , our study did not have sufficient statistical power ( statistical power range: 19–45% ) to replicate these markers ( Table 2 ) . Associations of T2D with SNPs that are in LD with the reported T2D SNPs discovered in European-ancestry populations or in Asians are presented in Table S3 . Multidimensional scaling analyses of the GWAS scan data showed no evidence of apparent genetic admixture in our study population ( Figure S2 ) . The observed number of SNPs with a small P value was larger than expected by chance ( Figure S3 ) . We found that rs10906115 ( 10p13 ) , rs1359790 ( 13q31 . 1 ) , and rs1436955 ( 15q22 . 2 ) were consistently associated with T2D across all studies , although the 95% CI for the per allele ORs in several studies included 1 . 0 ( Table 3; Figure 1 ) . P-values for trend tests ( per allele OR , 95% CI ) from meta-analyses of data from all studies were highly statistically significant for these associations: 1 . 45×10−8 for rs10906115 ( 1 . 13 , 1 . 08–1 . 18 ) , 6 . 49×10−9 for rs1359790 ( 1 . 15 , 1 . 10–1 . 20 ) , and 7 . 14×10−7 for rs1436955 ( 1 . 13 , 1 . 08–1 . 19 ) . These P-values were below ( for rs1359790 and rs10906115 ) or near ( for rs1436955 ) the genome-wide significance level of 5 . 0×10−8 . SNP rs10751301 ( 11q14 . 1 ) was not replicated in the Singapore or de novo genotyping studies; the P-value for the meta-analysis was 1 . 31×10−4 in the fixed effect model and 0 . 004 in the random effect model . Additional adjustment for smoking history did not appreciably change the point estimates described above , although the P-values were slightly elevated ( Table S2 ) . In an exploratory analysis stratified by smoking , BMI , family history of T2D , and age at diagnosis , SNP rs1359790 showed a slightly stronger association with T2D risk among non-smokers ( per allele OR = 1 . 19 , 95% CI = 1 . 12–1 . 26 , P = 6 . 4×10−8 ) than among smokers ( OR = 1 . 09 , 95% CI = 1 . 00–1 . 19 , P = 0 . 044 ) with a P value of 0 . 11 for interaction ( Table S4 ) . None of the SNPs were related to age at onset of T2D . Neither family history of T2D nor BMI altered the SNP-T2D associations under study .
Using the GWAS data from our discovery stage samples , we were able to validate 8 of 22 previously reported , GWAS-identified T2D SNPs , lending strong support to the validity of the initial discovery samples and methodologies . Applying a fast-track validation study approach , we also identified three promising new T2D markers . The most significant association identified by our study was for rs1359790 ( 13q13 . 1 ) , a novel genetic susceptibility locus identified for T2D ( Figure 2 ) . Several transcription factors , such as NIT-2 , CdxA , GATA-2 , and CDP , bind to this polymorphic site . The C to T transition eliminates a GATA-2 binding site and creates a TATA binding site . The closest known gene , sprouty homolog 2 ( Drosophila ) ( SPRY2 ) , is located 193 kb upstream of rs1359790 . The SPRY2 gene encodes a protein belonging to the sprouty family and inhibits growth factor-mediated , receptor tyrosine kinase-induced , mitogen-activated protein kinase signaling [13] . The encoded protein contains a carboxyl-terminal cysteine-rich domain essential for the inhibitory activity of receptor tyrosine kinase signaling proteins and is required for growth factor-stimulated translocation of the protein to membrane ruffles [13] , [14] . SPRY2 also modulates the apoptotic actions induced by the pro-inflammatory cytokine , tumor necrosis factor-alpha [15] . SPRY4 , a homolog of SPRY2 , inhibits the insulin receptor-transduced MAPK signaling pathway [16] and regulates development of the pancreas [17] . SNP rs10906115 is located on chromosome 10p13 ( Figure 2 ) , 13 . 0 kb from rs12779790 , which was reported by a previous GWAS of T2D [1] . These two SNPs , however , are in low LD in both Chinese ( r2 = 0 . 06 ) and European populations ( r2 = 0 . 19 ) based on HapMap data . SNP rs12779790 was not included in the Affymetrix SNP Array 6 . 0 , Illumina HumanHap 610-Quad , or Human1M-Duo; thus , it was imputed for both the SBCS/SWHS and the NHS/HPFS by using MACH with RSQR>0 . 9 and for the Singapore studies using IMPUTE with PROPER_INFO >0 . 85 . The imputed SNP rs12779790 was associated with a per allele OR of 1 . 10 ( 95% CI = 1 . 01–1 . 19 , P = 0 . 035 ) in the analysis of pooled data from three studies . However , when both rs12779790 and rs10906115 were included in the same logistic model , the association with rs10906115 remained statistically significant ( per allele OR = 1 . 09 , 95% CI = 1 . 02–1 . 16 , P = 0 . 007 ) , while the association with rs12779790 was no longer statistically significant ( per allele OR = 1 . 04 [95% CI = 0 . 96–1 . 12] , P = 0 . 38; Table 4 ) . These data provide strong evidence that rs10906115 is a new genetic variant at 10p13 independent of the previously-identified SNP rs12779790 . SNP rs10906115 is located 22 . 4 kb downstream of the cell division-cycle 123 homolog ( S . cerevisiae ) ( CDC123 ) gene and 76 . 6 kb upstream of the calcium/calmodulin-dependent protein kinase ID ( CAMK1D ) gene ( Figure 2 ) . The CDC123 gene encodes a protein involved in cell cycle regulation and nutritional control of gene transcription [18] . The CAMK1D gene encodes a member of the Ca2+/calmodulin-dependent protein kinase 1 subfamily of serine/threonine kinases . The encoded protein may be involved in the regulation of granulocyte function through the chemokine signal transduction pathway [19] . The role of the CDC123 and CAMK1D genes in the etiology of T2D is unclear . SNP rs1436955 , located on chromosome 15q22 . 2 ( Figure 2 ) , is 51 . 4 kb downstream of a C2 calcium-dependent domain containing the 4B gene ( C2CD4B; also known as NLF2 or FAM148B ) . C2CD4B is up-regulated by pro-inflammatory cytokines and may play a role in regulating genes that control cellular architecture [20] . The role of inflammation in the pathophsyiology of T2D has been suggested previously [21]–[25] . C2CD4B and SPRY2 are both highly expressed in human pancreatic tissue [26] . Intriguingly , a very recent report from the Meta-Analysis of Glucose and Insulin-related traits Consortium ( MAGIC ) found that a SNP ( rs11071657 ) near the C2CD4B gene was associated with fasting glucose ( P = 3 . 6×10−8 ) and T2D ( P = 2 . 9×10−3 ) [27] . SNPs rs11071657 and rs1436955 , however , are not in LD ( r2 = 0 . 04 ) in Asians , although they are weakly related ( r2 = 0 . 25 ) in Europeans , according to HapMap data . SNP rs11071657 is not included in the Affymetrix SNP 6 . 0 array . Imputed data from the SBCS/SWHS GWAS showed that this SNP was not significantly associated with T2D risk ( per A allele OR = 1 . 06 , 95% CI = 0 . 94–1 . 19 ) , although the direction of the association was consistent with that reported by the MAGIC consortium [27] . Adjusting for rs11071657 did not alter the association of T2D risk with rs1436955 ( per allele OR = 1 . 21 , 95% CI = 1 . 06–1 . 39 , P = 0 . 006 ) . Again , these data strongly imply that rs1436955 may be a new genetic risk variant for T2D at 15q22 . 2 independent of the recently reported SNP rs11071657 . In summary , in this first GWAS of T2D conducted in a Chinese population , we identified a novel genetic susceptibility locus for T2D , rs1359790 , at 13q31 . 1 . Furthermore , we revealed two new genetic variants ( rs10906115 at 10p13 and rs1436955 at15q22 . 2 ) near T2D susceptibility loci previously reported by GWAS of T2D conducted in European-ancestry populations . Our study demonstrates the value of conducting GWAS in non-European populations for the identification of novel genetic susceptibility markers for T2D . | Type 2 diabetes , a complex disease affecting more than a billion people worldwide , is believed to be caused by both environmental and genetic factors . Although some studies have shown that certain genes may make some people more susceptible to type 2 diabetes than others , the genes reported to date have only a small effect and account for a small proportion of type 2 diabetes cases . Furthermore , few of these studies have been conducted in Asian populations , although Asians are known to be more susceptible to insulin resistance than people living in Western countries , and incidence of type 2 diabetes has been increasing alarmingly in Asian countries . We conducted a multi-stage study involving 9 , 794 type 2 diabetes cases and 14 , 615 controls , predominantly Asians , to discover genes related to susceptibility to type 2 diabetes . We identified 3 genetic regions that are related to increased risk of type 2 diabetes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/gene",
"discovery",
"genetics",
"and",
"genomics"
] | 2010 | Identification of New Genetic Risk Variants for Type 2 Diabetes |
Studies in fission yeast have previously identified evolutionarily conserved shelterin and Stn1-Ten1 complexes , and established Rad3ATR/Tel1ATM-dependent phosphorylation of the shelterin subunit Ccq1 at Thr93 as the critical post-translational modification for telomerase recruitment to telomeres . Furthermore , shelterin subunits Poz1 , Rap1 and Taz1 have been identified as negative regulators of Thr93 phosphorylation and telomerase recruitment . However , it remained unclear how telomere maintenance is dynamically regulated during the cell cycle . Thus , we investigated how loss of Poz1 , Rap1 and Taz1 affects cell cycle regulation of Ccq1 Thr93 phosphorylation and telomere association of telomerase ( Trt1TERT ) , DNA polymerases , Replication Protein A ( RPA ) complex , Rad3ATR-Rad26ATRIP checkpoint kinase complex , Tel1ATM kinase , shelterin subunits ( Tpz1 , Ccq1 and Poz1 ) and Stn1 . We further investigated how telomere shortening , caused by trt1Δ or catalytically dead Trt1-D743A , affects cell cycle-regulated telomere association of telomerase and DNA polymerases . These analyses established that fission yeast shelterin maintains telomere length homeostasis by coordinating the differential arrival of leading ( Polε ) and lagging ( Polα ) strand DNA polymerases at telomeres to modulate Rad3ATR association , Ccq1 Thr93 phosphorylation and telomerase recruitment .
In eukaryotic cells , dynamic cell cycle-regulated protein-DNA complexes formed at telomeres play key roles in the maintenance of genome stability [1] , [2] . Telomeric DNA , consisting of repetitive GT-rich sequences , is extended by telomerase to overcome loss of telomeric DNA due to the inability of replicative DNA polymerases to fully replicate ends of linear DNA molecules [3] . While telomeric DNA is mostly double-stranded , telomeres terminate with a single-stranded GT-rich 3′ overhang , known as G-tail . Cells have evolved distinct proteins that specifically recognize either double-stranded or single-stranded telomeric DNA [4] . In mammalian cells , double-stranded DNA ( dsDNA ) -specific telomere binding proteins are encoded by TRF1 and TRF2 and a single-stranded DNA ( ssDNA ) -specific telomere binding protein is encoded by POT1 , and together with RAP1 , TIN2 and TPP1 , they form a telomere protection complex known as “shelterin” [4] . Mutations that affect shelterin or telomerase function in mammalian cells could lead to diseases that show premature aging due to depletion of the stem cell population , highlighting the importance to understand the regulatory mechanisms that ensure stable telomere maintenance [5] . Identification of a telomere protection complex that closely resembles mammalian shelterin [6] , coupled with the amenability to detailed genetic and molecular analysis , have made fission yeast Schizosaccharomyces pombe an attractive model organism to study telomere maintenance [7] . The shelterin complex in fission yeast consists of Taz1 ( TRF1/TRF2 ortholog ) that specifically recognizes double-stranded telomeres , the G-tail binding protein Pot1 , Tpz1 ( TPP1 ortholog ) , Rap1 , Poz1 and Ccq1 . In addition , Rif1 also interacts with Taz1 [8] . Similar to the way TIN2 and TPP1 connect TRF1/TRF2 to POT1 in mammalian shelterin , Rap1 , Poz1 and Tpz1 connect Taz1 to Pot1 ( Figure 1A ) . Ccq1 , which directly interacts with both Tpz1 and the telomerase regulatory subunit Est1 , plays a critical role in both recruitment of telomerase and attenuation of Rad3ATR-dependent DNA damage checkpoint responses [6] , [9] , [10] . Checkpoint kinases Rad3ATR and Tel1ATM are redundantly required for telomere maintenance and telomerase recruitment [11] , [12] , since the interaction between Ccq1 and the 14-3-3-like domain of Est1 is facilitated by Rad3ATR/Tel1ATM-dependent phosphorylation of Ccq1 on Thr93 [10] , [13] . Poz1 , Rap1 , and Taz1 are necessary to limit Ccq1 phosphorylation and uncontrolled telomere extension by telomerase [10] , but exactly how they prevent Rad3ATR/Tel1ATM-dependent phosphorylation of Ccq1 has not yet been established . In addition to shelterin , another evolutionarily conserved ssDNA binding complex , known as CST ( CTC1-STN1-TEN1 in mammalian cells and Cdc13-Stn1-Ten1 in budding yeast Saccharomyces cerevisiae ) , has been implicated in telomere maintenance [14]–[16] . CST interacts with the primase-DNA polymerase α complex [17]–[19] , and regulates G-tail length by promoting lagging strand synthesis at telomeres [20]–[22] . Furthermore , CST may inhibit telomerase activity by interacting with TPP1 [23] . While a Cdc13/CTC1-like protein has not yet been identified in fission yeast ( Figure 1A ) , deleting Stn1 or Ten1 resulted in immediate telomere fusion , highlighting the critical role of the Stn1-Ten1 complex in telomere maintenance [24] . Using Chromatin immunoprecipitation ( ChIP ) assays , we have previously established cell cycle-regulated changes in telomere association of telomere-specific proteins ( telomerase catalytic subunit Trt1TERT , Taz1 , Rap1 , Pot1 and Stn1 ) , DNA replication proteins ( DNA polymerases , MCM and RPA ) , the checkpoint protein Rad26ATRIP ( a regulatory subunit of checkpoint kinase Rad3ATR ) and DNA repair protein Nbs1 ( a subunit of Mre11-Rad50-Nbs1 complex ) in fission yeast [25] . Unexpectedly , the leading strand DNA polymerase Polε arrived at telomeres significantly earlier than the lagging strand DNA polymerases Polα and Polδ in late S-phase . Temporal recruitment of RPA and Rad26ATRIP matched the arrival of Polε , while recruitment of Trt1TERT , Pot1 and Stn1 matched the arrival of Polα . However , it has not yet been established if the delayed arrival of Polα/Polδ represents a C-strand fill-in reaction after extension of the G-strand by telomerase , or if it might be part of the regulatory mechanism that controls recruitment of telomerase by regulating Rad3ATR/Tel1ATM accumulation and Ccq1 Thr93 phosphorylation . While previous studies have established that Taz1 and mammalian TRF1 contribute to efficient replication of telomeric repeats [26] , [27] , very little is known how the loss of Taz1 or TRF1 affects behaviors of replicative DNA polymerases at telomeres . In addition , it is currently unknown how cell cycle-regulated dynamic binding patterns of checkpoint kinases , shelterin and CST are affected by challenges posed by replicating highly extended telomeric repeats as found in poz1Δ , rap1Δ , and taz1Δ cells . Therefore , we investigated how loss of the shelterin subunits Poz1 , Rap1 and Taz1 affects cell cycle-regulated recruitment timing of telomerase catalytic subunit Trt1TERT , DNA polymerases ( Polα and Polε ) , the Replication Protein A ( RPA ) complex subunit Rad11 , the Rad3ATR-Rad26ATRIP checkpoint kinase complex , Tel1ATM kinase , shelterin subunits ( Tpz1 Ccq1 and Poz1 ) , and Stn1 . In addition , we investigated how telomere shortening , caused either by deletion of Trt1TERT or introduction of catalytically dead Trt1TERT , affected cell cycle-regulated telomere association of telomerase and DNA polymerases . Our detailed ChIP analyses provide new insights into the dynamic coordination of DNA replication , DNA damage kinase recruitment , and telomerase recruitment in fission yeast .
To better understand how Poz1 , Rap1 and Taz1 function together in telomere maintenance , we performed epistasis analysis among single , double and triple deletion mutant cells for telomere length , cold sensitivity , protection of telomeres against telomere fusion in G1 arrested cells , and recruitment of Trt1TERT to telomeres [6] , [8] , [28] , [29] . Telomere length distribution of poz1Δ , rap1Δ and poz1Δ rap1Δ cells closely resembled one another ( Figures S1A and 1B #1 ) , suggesting that poz1Δ and rap1Δ cause similar defect ( s ) in telomere length regulation . The distribution of telomere length was broader and skewed toward shorter telomeres in taz1Δ cells than rap1Δ or poz1Δ cells ( Figure 1B #2-3 ) , and rap1Δ taz1Δ and poz1Δ rap1Δ taz1Δ cells showed identical telomere length distributions as taz1Δ cells ( Figure 1B #4 ) , suggesting that Taz1 carries out both Poz1/Rap1-dependent and -independent roles in telomere length regulation . Interestingly , since telomere length distribution in poz1Δ taz1Δ was much broader than in poz1Δ rap1Δ taz1Δ cells ( Figure 1B #3-4 ) , it appears that Rap1 could also affect telomere length independently of Poz1 and Taz1 . In support for such independent function , Rap1 binding to telomeres was significantly reduced but not entirely eliminated in poz1Δ taz1Δ cells ( Figure 1C ) . Previously , we have also found that Rap1 contributes to recombination-based telomere maintenance independently of Taz1 and Poz1 [30] . We also found that poz1Δ and taz1Δ cells , but not rap1Δ cells , show reduced cell growth at lower temperature ( Figure S1B ) . Cold sensitivity of taz1Δ cells [28] was more severe than poz1Δ cells , and poz1Δ taz1Δ cells were more sensitive than taz1Δ cells . Interestingly , while rap1Δ taz1Δ cells showed the most severe cold sensitivity among all mutant combinations tested , cold sensitivity of poz1Δ rap1Δ taz1Δ cells was milder , suggesting that the presence of Poz1 in rap1Δ taz1Δ cells is detrimental to cell growth at low temperature . In addition , rap1Δ and taz1Δ cells , but not poz1Δ cells , showed telomere-telomere fusion [28] , [31] , [32] when cells are grown in low nitrogen media to arrest cells in G1 ( Figure S1C ) . Among double and triple mutant cells , all cells that lack Rap1 and/or Taz1 underwent telomere fusion . Thus , only Rap1 and Taz1 ( but not Poz1 ) are involved in protection of telomeres against fusions in G1 arrested cells . Based on ChIP analysis utilizing the hybridization of a telomeric probe to dot blotted samples , we found that Trt1TERT showed progressive increase in telomere association in the order of poz1Δ , rap1Δ and taz1Δ cells [10] ( Figure 1D ) . Further analysis of double and triple mutant cells revealed that poz1Δ rap1Δ cells have similar levels of Trt1TERT binding as rap1Δ cells , and poz1Δ taz1Δ , rap1Δ taz1Δ and poz1Δ rap1Δ taz1Δ cells have similar levels of Trt1TERT binding as taz1Δ cells . Thus , regarding its inability to limit telomerase binding to telomeres , taz1Δ is epistatic over rap1Δ or poz1Δ , and rap1Δ is epistatic over poz1Δ . Trt1TERT binding to telomeres is cell cycle-regulated , and maximal association of Trt1TERT occurs in late-S phase when telomeres are replicated [25] . To better understand the roles of Poz1 , Rap1 and Taz1 in limiting Trt1TERT binding to telomeres , we examined changes in Trt1TERT association in poz1Δ , rap1Δ and taz1Δ cells by ChIP using cdc25-22 synchronized cell cultures . Since our asynchronous ChIP analysis indicated that Trt1TERT recruitment to telomeres is similar in single and double/triple mutant cells , we limited our cell cycle ChIP analysis to single mutant cells . After incubating cdc25-22 cells at non-permissive temperature ( 36°C ) for 3 hours , late G2-phase arrested cells were shifted to permissive temperature ( 25°C ) for synchronous cell cycle re-entry , and samples were collected every 20 minutes and processed for ChIP analysis . In previous cell cycle ChIP analyses , we utilized quantitative real-time PCR with primers that amplify a unique sub-telomeric DNA sequence directly adjacent to telomeric repeats [12] , [25] . Since wild-type ( wt ) telomeres are only ∼300 bp and the size of DNA fragments after sonication is estimated to be in the 0 . 5∼1 kb range , the use of sub-telomeric primers provides a convenient mean to determine the association of various factors to telomeres . However , since poz1Δ , rap1Δ and taz1Δ cells carry much longer telomeres , sub-telomeric PCR primer pairs would be too distant from the actual chromosome ends , and thus protein binding to telomeres had to be monitored using dot blotted samples and utilizing telomeric repeat DNA as hybridization probe [10] . For analysis of dot blot-based ChIP assays , we processed raw data ( % precipitated DNA ) in two different ways . First , to compare overall temporal binding patterns , we normalized ChIP data to the peak of binding within the first complete cell cycle ( 40–200 min ) after release from the G2 arrest . Second , we attempted to obtain an approximate fold-increase in protein association “per chromosome end” by correcting for changes in telomeric tract length ( Materials and Methods and Table S1 ) . This correction was necessary as raw % precipitated DNA values reflect the density of a given protein within the telomeric tract , and thus significantly underestimate the actual increase in protein binding at chromosome ends for cells carrying long telomeric repeat tracts . Telomere length corrected ChIP data were normalized to values from wt cells for asynchronous ChIP assays , and normalized to the peak binding values of wt cells in late S/G2-phase for cell cycle ChIP assays . ( See Figures S2 for telomere length correction of Trt1TERT asynchronous ChIP data as example . ) Based on changes in % septated cells , poz1Δ , rap1Δ and taz1Δ cells showed similar re-entries into cell cycle as wt cells ( Figure S3C ) , with the first S-phase occurring 60–140 min and the second S-phase starting 200–220 min after the temperature shift . BrdU incorporation data indicated that telomeres in wt , poz1Δ and rap1Δ cells are replicated in late S-phase ( 100–140 min after the temperature shift ) , while replication of telomeres in taz1Δ cells occurred much earlier ( 60–100 min after the temperature shift ) ( Figure S4B ) . Furthermore , hydroxyurea ( HU ) treatment completely abolished telomere replication in wt , poz1Δ and rap1Δ cells , but not in taz1Δ cells . These data are consistent with previous findings that Taz1 is required to enforce late S-phase replication at telomeres [33] , [34] . Consistent with our previous analysis [25] , Trt1TERT showed maximal binding to telomeres in late S-phase ( 120–140 min ) in wt cells ( Figure 2A ) . In poz1Δ and rap1Δ cells , Trt1TERT showed nearly identical cell cycle-regulated association patterns with a substantial delay in maximal binding ( 160–180 min ) ( Figure 2A ) . In agreement with a recent report [34] , we found that Trt1TERT is bound to telomeres throughout the cell cycle in taz1Δ cells with much broader and persistent maximal binding at 120–180 min ( Figures 2B and S3A–B ) . Consistent with asynchronous ChIP data , relative peak binding values ( telomere length corrected ) for Trt1TERT increased in the order of poz1Δ ( ∼40-fold ) , rap1Δ ( ∼59-fold ) and taz1Δ ( ∼167-fold ) over wt cells ( Figure 2B ) . Real-time PCR-based ChIP assays have previously established that the leading strand DNA polymerase Polε arrives at telomeres significantly earlier than the lagging strand DNA polymerases Polα and Polδ , and that the timing of maximal Trt1TERT association matches more closely to that of Polα and Polδ ( ∼140 min ) than Polε ( ∼120 min ) [25] . Our dot blot-based ChIP re-confirmed the differential timing in peak association for Polα and Polε in wt cells ( Figures 2C and S5 ) . In poz1Δ and rap1Δ cells , binding of Polα was delayed ∼40 min without affecting the temporal binding pattern of Polε . The delay of Polα appears to be restricted to telomeres , as the timing of Polα association with ars2004 ( early replication origin ) was similar among wt , poz1Δ and rap1Δ cells ( Figure S4C ) . Overall , the cell cycle-regulated association patterns for both polymerases were nearly identical in poz1Δ and rap1Δ cells , but both Polα and Polε showed increased association with telomeres in poz1Δ cells than rap1Δ cells ( Figures 2C and S5A–B ) . In taz1Δ cells , the difference in telomere binding patterns for the leading and lagging strand DNA polymerases was more dramatic . As expected based on the fact that taz1Δ cells replicate telomeres much earlier in S-phase [33] ( Figure S4B ) , Polε was recruited to telomeres earlier ( peak binding ∼100 min ) ( Figure S5B ) . When corrected for telomere length , we found a ∼6 fold increase in peak ChIP precipitation for Polε in taz1Δ cells over wt cells ( Figure 2C ) . Surprisingly , Polα was constitutively bound to telomeres throughout the cell cycle in taz1Δ cells at ∼1 . 5 fold above the peak binding in wt cells ( Figures 2C and S5A ) . On the other hand , overall cell cycle progression ( Figure S5E–F ) and association timing for Polα to ars2004 ( Figure S4C ) were not affected in taz1Δ cells . Taken together , we concluded that Poz1 and Rap1 are required primarily to maintain timely recruitment of Polα to telomeres , and Taz1 is required to both ( 1 ) delay arrival of Polε to enforce late S-phase replication of telomeres and ( 2 ) enforce cell cycle-regulated association of Polα with telomeres . Previous ChIP analysis using real-time PCR found largely overlapping temporal association patterns for the telomerase catalytic subunit Trt1TERT and Polα with both showing maximal binding at ∼140 min in wt cells [25] . However , the initial increase in detectable binding to telomeres was earlier for Trt1TERT ( ∼80 min ) than Polα ( ∼100 min ) and treatment with HU caused much greater inhibition of Polα and Polε binding than Trt1TERT , suggesting that Trt1TERT binding could occur prior to the arrival of replicative polymerases at telomeres [25] . With dot blot-based ChIP analysis , the overall binding pattern for Trt1TERT was broader than in our previous analysis ( Figure 2A ) [25] . Thus , when data for Trt1TERT , Polα and Polε were plotted together ( Figure 2D ) , the increase in Trt1TERT binding prior to arrival of Polα became more evident . On the other hand , reductions in the binding of Trt1TERT and Polα in G2/M phase occurred with very similar timing . In poz1Δ and rap1Δ cells , the peak of Trt1TERT recruitment was dramatically delayed compared to Polε and its overall temporal association pattern largely overlapped with Polα ( Figure 2D ) . However , the initial increase in Trt1TERT binding to telomeres occurred with similar timing as Polε in poz1Δ , rap1Δ or taz1Δ cells ( Figure S6A ) , and the amount of Trt1TERT binding was already significantly increased in early S-phase ( 80–100 min ) and further elevated during late S/G2-phases ( 160–180 min ) in these deletion mutants ( Figure 2B ) . Thus , the delay in peak binding of Trt1TERT in poz1Δ and rapΔ cells is caused primarily by the massive increase in Trt1TERT binding during late S/G2-phases . Likewise , the broad and persistent binding of Trt1TERT in taz1Δ cells can be attributed to both a massive increase in early S-phase and persistent binding in late S/G2-phases . Taken together , we thus concluded that Trt1TERT binding to telomeres occurs around the time when Polε arrives at telomeres , and that its binding is massively increased throughout S-phase in cells that lack Poz1 , Rap1 or Taz1 , accompanied by delayed ( poz1Δ and rap1Δ ) or persistent ( taz1Δ ) binding of Polα . The differential arrival of leading and lagging strand DNA polymerases could temporarily create extended ssDNA at telomeres that are then replicated by the lagging strand polymerase . Indeed , both the largest subunit of the ssDNA binding complex RPA ( Rad11 ) and the checkpoint kinase regulatory subunit Rad26 ( ATRIP ortholog ) showed increased binding to telomeres as the leading strand DNA polymerase ( Polε ) arrives and reduced binding as the lagging strand DNA polymerases ( Polα and Polδ ) arrive at telomeres [25] ( Figure 3C wt ) . Since Polα association is even more delayed in poz1Δ and rap1Δ cells and severely deregulated in taz1Δ cells ( Figure 2C–D ) , we predicted that both the Rad3ATR-Rad26ATRIP complex and Rad11RPA to increase in telomere association upon loss of Poz1 , Rap1 or Taz1 . Indeed , asynchronous ChIP assays found that Rad3ATR , Rad26ATRIP and Rad11RPA all show a significant increase in binding to telomeres in poz1Δ , rap1Δ and taz1Δ cells ( Figures S7A and S8 ) . The extent of increase was much greater for Trt1TERT and Rad26ATRIP than Rad3ATR and Rad11RPA , but all showed a much greater degree of binding increase to telomeres than shelterin subunits ( Tpz1 , Ccq1 and Poz1 ) or Stn1 ( Figures S7B and S9 ) . In contrast to Rad3ATR-Rad26ATRIP , Tel1ATM kinase did not show much increase in telomere association upon elimination of Poz1 , Rap1 or Taz1 , even though Rad3ATR and Tel1ATM play redundant role ( s ) in telomere protection and telomerase recruitment ( Figure S10 ) . These data are consistent with previous conclusions that Rad3ATR plays a much greater role in regulation of telomere length and Ccq1 phosphorylation than Tel1ATM in fission yeast [10] , [12] , [13] . While Rad26ATRIP and Rad11RPA association increased throughout the cell cycle in poz1Δ and rap1Δ cells compared to wt , the most noticeable change was their increased and persistent binding during the extended time period ( 80–200 min ) between the arrival of Polε and dissociation of Polα ( Figures 3 and S11A–B ) . While increases in telomere binding during S-phase were more dramatic for Rad26ATRIP than Rad11RPA ( Figure 3B ) , both proteins showed significantly higher binding to telomeres in rap1Δ than in poz1Δ cells ( Figure 3A ) , consistent with asynchronous ChIP data ( Figure S7A ) and our previous findings that rap1Δ cells show stronger induction of Ccq1 Thr93 phosphorylation and increased binding of Trt1TERT than poz1Δ cells [10] . For taz1Δ cells , both Rad26ATRIP and Rad11RPA showed their strongest binding to telomeres immediately after release from cdc25-22 induced G2 arrest ( Figures 3A and S11A–D ) , suggesting that prolonged arrest in G2 might cause continued resection of telomeric ends and much higher levels of Rad3ATR-Rad26ATRIP and Rad11RPA accumulation specifically in taz1Δ cells . Nevertheless , both Rad26ATRIP and Rad11RPA showed significant reduction in telomere association as cells completed mitosis ( ∼80 min ) , increased and persistent binding during S/G2-phase , and slight reduction in binding in late G2/M-phase ( Figures 3 and S11A–D ) . Thus , despite the lack of any observable cell cycle regulation for Polα association with telomeres in taz1Δ cells , there must be some changes at taz1Δ telomeres that allow a slight reduction in association of the Rad3ATR-Rad26ATRIP kinase complex and RPA in late G2/M-phase . Phosphorylation of Ccq1 Thr93 by Rad3ATR and Tel1ATM kinases is important for telomerase recruitment in fission yeast [10] , [13] . Since Ccq1 is hyper-phosphorylated in poz1Δ , rap1Δ , or taz1Δ cells at Thr93 and additional unidentified phosphorylation sites [10] , we next examined how Ccq1 phosphorylation is regulated during cell cycle . While massively increased in rap1Δ and taz1Δ over wt cells , the overall phosphorylation status of Ccq1 , monitored by the presence of a slow mobility band of Ccq1 on SDS-PAGE ( marked with * ) , was constant and did not show any cell cycle regulation in all genetic backgrounds tested ( Figure 4A ) . In contrast , Thr93-dependent phosphorylation of Ccq1 , detected by phospho- ( Ser/Thr ) ATM/ATR substrate antibody [10] ( see comment in Materials and Methods ) , showed cell cycle-regulated changes . In wt cells , Thr93 phosphorylation peaked during late S-phase ( 100–140 min ) , but was quickly reduced at later time points and nearly abolished at 200 min before cells entered their next S-phase ( Figure 4A ) . Thus , Thr93 phosphorylation was reduced with similar timing as Trt1TERT ( Figure 2A–B ) and Rad26ATRIP ( Figure S11A ) binding at 160–200 min . In rap1Δ and taz1Δ cells , Thr93 phosphorylation was increased throughout the entire cell cycle with slight reductions at 60 and 180–200 min ( Figure 4A ) , but did not entirely match the temporal recruitment pattern of Trt1TERT to telomeres , which showed a dramatic increase in binding in late S-phase . Thus , we concluded that there must be other cell cycle-regulated changes besides Ccq1 Thr93 phosphorylation that regulate Trt1TERT recruitment to telomeres . Previous ChIP analysis had revealed that the shelterin ssDNA-binding subunit Pot1 along with the CST-complex subunit Stn1 show significant late S-phase specific increases in telomere association that matched to the timing of Polα and Trt1TERT recruitment [25] . We reasoned that cell cycle-regulated changes in shelterin and CST telomere association could dictate Trt1TERT binding , and thus decided to monitor how loss of Poz1 , Rap1 and Taz1 affect cell cycle-regulated association of shelterin and CST . We limited our analysis to three subunits of shelterin ( Ccq1 , Tpz1 and Poz1 ) and Stn1 , and decided to exclude Pot1 , since we found that addition of an epitope tag to Pot1 significantly altered telomere length of poz1Δ , rap1Δ and taz1Δ cells . Consistent with asynchronous ChIP data ( Figure S7B ) , Ccq1 , Tpz1 , Poz1 and Stn1 all showed gradual increases in overall binding to telomeres in the order of wt , poz1Δ , rap1Δ and taz1Δ when corrected for changes in telomere length ( Figure 4B ) . Ccq1 and Tpz1 showed nearly identical temporal recruitment patterns in wt , poz1Δ , rap1Δ , and taz1Δ cells ( Figure S13 ) , while Poz1 recruitment was delayed compared to Ccq1 and Tpz1 , and more closely resembled the pattern found for Stn1 ( Figures S14 and S15 ) . We were initially surprised by the similarity of the temporal recruitment patterns for Poz1 and Stn1 , as we previously failed to detect interaction between shelterin and Stn1-Ten1 by co-immunoprecipitation [25] . On the other hand , studies in mammalian cells have detected TPP1-CST interaction [23] , [35] , and we also found by 3-hybrid assay that Tpz1 can interact with Stn1-Ten1 ( Figures 4C and S16 ) . Intriguingly , the Tpz1 interaction with Stn1-Ten1 became stronger when the Ccq1/Poz1 interaction domain of Tpz1 ( amino acids 421–508 ) was deleted , suggesting that this domain might negatively regulate the interaction between Tpz1 and Stn1-Ten1 . Thus , it is possible that Tpz1-Poz1 interaction might facilitate the timely recruitment of Stn1-Ten1 by reducing the ability of the Tpz1 C-terminal domain to negatively regulate interaction between Tpz1 and Stn1-Ten1 . Comparison with DNA polymerases revealed that Ccq1 and Tpz1 show increases in telomere association along with Polε ( 80–120 min ) and reduction in binding along with Polα ( 140–220 min ) in wt cells ( Figure S17 ) . The onsets of increased binding in Ccq1 and Tpz1 remained similar ( ∼80 min ) in the deletion mutants . However , Ccq1 and Tpz1 binding peaked at 140 min , between the peaks for Polε and Polα in poz1Δ and rap1Δ cells , while they sustained increased binding longer ( 120–180 min ) in taz1Δ ( Figures 4B and S17 ) . Thus , analogous to Rad26ATRIP ( Figure 3 ) , increased binding of Ccq1 and Tpz1 during S-phase in poz1Δ , rap1Δ and taz1Δ cells may be dictated by increased ssDNA caused by deregulated replication of telomeres . In contrast , the temporal binding patterns for Stn1 and Poz1 matched closely with the binding pattern for Polα ( Figure 5A ) in all genetic backgrounds tested , except for taz1Δ . This is consistent with the notion that Poz1 and Stn1 may closely collaborate in promoting the timely recruitment of Polα to telomeres . We also found that Stn1 in wt , poz1Δ and rap1Δ cells shows more persistent binding at later time points than Polα ( Figure 5A ) , suggesting that Stn1 can sustain increased telomere binding even after Polα dissociates from telomeres . Consistently , we have previously observed increased binding of Stn1 to telomeres in S-phase , even when Polα recruitment was inhibited by HU treatment [25] . Ccq1 , Tpz1 and Trt1TERT showed nearly identical overall temporal binding patterns in wt and taz1Δ cells , consistent with the notion that cell cycle-regulated binding of Tpz1 and Ccq1 plays a major role in controlling Trt1TERT association with telomeres ( Figure 5B ) . In contrast , Trt1TERT reached its maximal binding later than Ccq1 and Tpz1 in poz1Δ and rap1Δ cells ( Figure 5B ) . However , this delay is a reflection of the dramatic increase in Trt1TERT binding at 160–200 min in poz1Δ and rap1Δ cells ( Figure 2A–B ) , a time period in which Ccq1 Thr93 phosphorylation is rapidly reduced in wt cells but remained constitutively high in rap1Δ or taz1Δ cells ( Figure 4A ) . Indeed , while huge increases in Trt1TERT binding over Tpz1 or Ccq1 had made it difficult to compare cell cycle-regulated patterns in linear scale plots , plotting data on log scale made it more clear that the initial increase in binding of Trt1TERT , Ccq1 and Tpz1 occurred with similar timing even in poz1Δ and rap1Δ cells ( Figure 5C ) . Poz1 and Stn1 binding to telomeres was delayed compared to Trt1TERT in wt cells , but all three proteins showed very similar overall temporal binding patterns in deletion mutant cells except for more persistent Stn1 binding at later time points ( Figure S18A–B ) . However , since Trt1TERT binding in poz1Δ , rap1Δ and taz1Δ cells increased even in early S-phase , the initial increase in Trt1TERT binding still preceded binding increases of Poz1 and Stn1 in deletion mutant backgrounds ( Figure S18C ) . Taken together , our findings are consistent with the notion that the initial increase in binding of Tpz1 and Ccq1 to telomeres in S-phase contributes to Trt1TERT recruitment , and that a subsequent increase in binding of Poz1 and Stn1 contributes to the timely recruitment of Polα , which limits ssDNA and Rad3ATR-Rad26ATRIP accumulation , Ccq1 Thr93 phosphorylation , and telomerase binding at telomeres . Ccq1 Thr93 phosphorylation is also increased in cells carrying short telomeres [10] , [13] . As short telomeres would have less binding sites for Taz1 [36] , [37] , they may become less effective in excluding the Rad3ATR-Rad26ATRIP complex from telomeres . Consistently , we found that Rad26ATRIP binding is indeed significantly increased in trt1Δ cells ( Figure 6A ) . While the notion that telomerase is preferentially recruited to short telomeres , due to reduced binding of Taz1 and increased Ccq1 Thr93 phosphorylation , is an attractive model to explain telomere length homeostasis in fission yeast , there has been a lack of any direct evidence that Trt1TERT binding is indeed increased at short telomeres [10] . The problem was difficult to address since mutations previously used to induce telomere shortening ( trt1Δ or ccq1-T93A ) eliminated telomerase or its recruitment [10] . We overcame this limitation by monitoring telomere binding of catalytically inactive Trt1TERT ( trt1-D743A ) , which causes telomere shortening [38] . Consistent with the prediction , we found that Trt1-D743A binds stronger than wt Trt1TERT to telomeres in asynchronous cell cultures ( Figure 6B and S19B ) , and binds constitutively throughout the cell cycle with increase in binding during S/G2-phase ( Figure 6C ) . Deletion of Rap1 further increased Trt1-D743A binding ( Figure 6B–C ) , especially at the early time points after cell cycle re-entry ( 20–60 min ) , but did not greatly affect the temporal recruitment pattern of Trt1-D743A for the remainder of the cell cycle . We are not entirely sure why trt1-D743A rap1Δ shows increased Trt1TERT binding at early time points , but binding levels were comparable to wt Trt1TERT in rap1Δ cells at 20–80 min ( Figure 6C ) . In contrast , far more wt Trt1TERT was recruited to telomeres than Trt1-D743A for the remainder of the cell cycle in rap1Δ cells , suggesting that the massive increase in telomere association of Trt1TERT in rap1Δ cells during late S/G2-phase is largely dependent on telomerase activity ( Figure 6C ) . We next investigated how DNA polymerases were affected in trt1Δ or trt1-D743A cells . Interestingly , Polε binding to telomeres peaked slightly earlier , and the initial increase in Polα binding also occurred slightly earlier in trt1Δ and trt1-D743A ( ∼100 min ) than wt cells ( ∼120 min ) ( Figure 7A ) . In contrast , we did not see any major change in Polα recruitment timing at ars2004 in trt1Δ cells ( Figure S20E ) . Since taz1Δ cells show earlier recruitment of Polε and telomere replication [33] ( Figures S4 and S5 ) , it is tempting to speculate that reduced binding of Taz1 at short telomeres might be responsible for earlier Polε recruitment ( consistent with earlier telomere replication ) for cells carrying short telomeres . While there was no obvious difference between trt1Δ and trt1-D743A cells in overall temporal recruitment patterns of Polε and Polα ( Figure 7A ) , when normalized for telomere length , trt1-D743A cells had significantly reduced binding of Polα compared to trt1Δ cells , suggesting that the presence of catalytically inactive Trt1TERT may interfere with efficient recruitment of Polα ( Figure 7B ) . Our data also indicated that Polε still arrives at telomeres significantly earlier than Polα in trt1Δ or trt1-D743A cells ( Figure 7C ) , suggesting that telomerase-dependent telomere extension cannot solely be responsible for the differential arrival of Polε and Polα at telomeres . By examining the temporal telomere association patterns of DNA polymerases in rap1Δ trt1Δ cells , we attempted to investigate if the delay of Polα arrival at telomeres in rap1Δ cells ( Figure 2C–D ) is dependent on telomerase . To our surprise , rap1Δ trt1Δ cells showed very little cell cycle-regulated Polα recruitment to telomeres ( Figure 8A ) , suggesting that Trt1TERT and Rap1 might play redundant roles in coordinating the lagging strand DNA synthesis at telomeres . However , since cells carrying Pol1-FLAG progressed substantially faster through the cell cycle in trt1Δ rap1Δ than wt cells ( Figure S21D ) , epitope-tagging of Polα may have introduced unintended changes in telomere regulation that caused synergistic genetic interactions specifically in rap1Δ trt1Δ cells . In contrast , we did not see much change in the temporal association pattern of Polε or cell cycle progression between wt and rap1Δ trt1Δ for cells carrying Pol2-FLAG ( Figures 8B and S21E ) . Because studies in other organisms have implicated a connection between Polα and CST in telomere regulation [17]–[19] and our cell cycle ChIP data revealed very similar timings of telomere association for Polα and Stn1 ( Figure 5A ) , we next examined the cell cycle-regulated association of Stn1 in rap1Δ trt1Δ cells . Much like Polα , S phase-induced increase in telomere binding of Stn1 was abolished in rap1Δ trt1Δ cells ( Figure 8C ) . However , we also noticed that Stn1-myc cells progressed through cell cycle slower in rap1Δ trt1Δ ( Figure S21F ) . Thus , epitope-tagging of Stn1 may have elicited unexplained additional telomere defects in rap1Δ trt1Δ cells . In any case , it was striking to find loss of cell cycle-regulated binding for both Polα and Stn1 without affecting Polε association in rap1Δ trt1Δ cells , and it might indicate that Rap1 and Trt1 play unexpected redundant roles in maintaining proper cell cycle-regulated localization of both Polα and Stn1-Ten1 to telomeres . It is worth noting that a recent study has found that inhibition of telomerase leads to reduced recruitment of Stn1 to telomeres in late S/G2-phase in mammalian cells , suggesting that mammalian telomerase also contributes to efficient recruitment of the CST complex to telomeres [23] .
While previous studies have implicated Taz1 and TRF1 in efficient replication of telomeric DNA [26] , [27] , very little was known how loss of Taz1/TRF1 affects replicative DNA polymerases at telomeres . We found that loss of telomerase inhibitors ( Poz1 , Rap1 and Taz1 ) differentially affect leading ( Polε ) and lagging ( Polα ) strand DNA polymerases ( Figure 2C ) . For poz1Δ and rap1Δ cells , the peak of Polα binding to telomeres was significantly delayed without affecting Polε , suggesting that Poz1 and Rap1 primarily affect the timely recruitment of the lagging strand DNA polymerase . Consistent with previous studies that observed more severe defects in telomere replication in taz1Δ than rap1Δ cells [28] , [34] , [41] , Polα binding to telomeres was severely deregulated in taz1Δ cells . In addition , loss of Taz1 ( but not Rap1 or Poz1 ) caused earlier recruitment of Polε to telomeres , consistent with recent findings that Taz1 and Taz1-interacting protein Rif1 enforce late S-phase replication of telomeres in fission yeast [33] , [42] . Intriguingly , telomerase deficient cells ( trt1Δ or trt1-D743A ) , which carry shorter telomeres and thus can accommodate less Taz1 , also showed slightly earlier recruitment of Polε to telomeres than wt cells , consistent with earlier replication of telomeres ( Figure 7 ) . Taken together , we thus propose that ( 1 ) Taz1 , likely in collaboration with Rif1 but independently of Poz1 and Rap1 , enforces late S-phase replication of telomeres , and as a consequence , ( 2 ) shorter telomeres in fission yeast are replicated earlier ( Figure 9 ) . Previously , we have found that Taz1 binding is reduced by ∼2-fold during S-phase [25] . Therefore , we speculate that shorter telomeres may be able to reduce Taz1 ( and Rif1 ) density faster to the level compatible with replication , and as a consequence , replicate earlier in S-phase than longer telomeres . Interestingly , budding yeast cells also replicate shorter telomeres earlier [43] , suggesting that early replication of short telomeres may be evolutionarily conserved . Moreover , even though replication timing of mammalian cells appears to be not strictly dependent on telomere length [44] , TRF1 binding to telomeres is also reduced during S-phase [45] and Rif1 , like fission yeast Rif1 , contributes to genome-wide regulation of replication timing [46] , [47] . It is thus possible that TRF1 might also collaborate with Rif1 in regulating replication timing at telomeres in mammalian cells . Epistasis analysis of telomere length by Southern blot indicated that Taz1 carries out both Poz1/Rap1-independent and -dependent roles in regulation of telomere length maintenance ( Figure 1B ) . Accordingly , we further suggest that Taz1 , in collaboration with Rap1 and Poz1 , also contributes to replication fork integrity at telomeres by promoting timely and cell cycle-regulated recruitment of the lagging strand DNA polymerases ( Polα and Polδ ) ( Figure 9 ) . Completion of the C-strand fill-in synthesis by lagging strand DNA polymerases would then restore dsDNA and Taz1 binding in late S/G2 phase , prior to the initiation of mitosis . Although Rap1 binding to telomeres is not entirely dependent on Taz1 , loss of Taz1 significantly reduces Rap1 binding at telomeres ( Figure 1C ) , and disruption of the Taz1-Rap1 interaction causes massive elongation of telomeres , much like rap1Δ [40] . Thus , it is easy to imagine how Taz1 , through direct interaction with Rap1 , would affect Rap1-dependent promotion of lagging strand synthesis at telomeres . Previously identified Rap1-Poz1 and Poz1-Tpz1 interactions [6] would also likely be important in regulating the timely recruitment of lagging strand DNA polymerases ( Figure 1A and Figure 9 ) . In addition , a newly identified interaction between Tpz1 and Stn1-Ten1 ( Figure 4C ) could play a critical role in allowing Rap1 and Poz1 to enforce the timely recruitment of Polα to telomeres , since our ChIP assays also implicated a close functional relationship among Poz1 , Stn1 and Polα in telomere regulation ( Figures 5A and 8 ) . Our current findings are likely relevant to understand mammalian telomere regulation , since previous studies have found that mammalian TPP1 also interacts with CST [23] , [35] , and CST collaborates with Polα in regulating C-strand fill-in synthesis at telomeres [20]–[22] . Since Rad3ATR/Tel1ATM-dependent phosphorylation of Ccq1 at Thr93 promotes Ccq1-Est1 interaction and telomerase recruitment , the mechanism that modulates Thr93 phosphorylation is critical for proper maintenance of telomeres in fission yeast [10] , [13] . Poz1 , Rap1 and Taz1 negatively regulate Ccq1 Thr93 phosphorylation and telomerase recruitment [10] , but the underlying mechanism by which these factors limit Thr93 phosphorylation remained unclear . Our ChIP data indicated that loss of Poz1 , Rap1 and Taz1 causes large increases in telomere association for RPA and Rad3ATR-Rad26ATRIP , but not Tel1ATM kinase ( Figures 3 , S7A and S10 ) . Closely matching the extent of increase in Ccq1 Thr93 phosphorylation and Trt1TERT binding [12] , RPA and Rad3ATR-Rad26ATRIP showed progressive increase in telomere binding in the order of poz1Δ , rap1Δ and taz1Δ cells , especially during S-phase . Therefore , we suggest that Poz1 , Rap1 and Taz1 negatively regulate Rad3ATR-Rad26ATRIP accumulation and Ccq1 Thr93 phosphorylation by controlling the differential arrival of leading and lagging strand polymerases at telomeres ( Figure 9 ) . Based on our cell cycle analysis , we further suggest that S-phase specific Trt1TERT recruitment to telomeres is controlled by both ( 1 ) cell cycle-regulated binding of Pot1-Tpz1-Ccq1 and ( 2 ) Ccq1 Thr93 phosphorylation . Since Thr93 phosphorylation is quickly lost in wt cells soon after dissociation of Rad26ATRIP from telomeres , it is likely that an unidentified phosphatase is involved in rapidly reducing Thr93 phosphorylation to promote the timely dissociation of Trt1TERT from telomeres . In poz1Δ , rap1Δ and taz1Δ cells , increased accumulation of Rad3ATR kinase results in constitutive Thr93 phosphorylation , hence persistent and high level binding of Trt1TERT in G2 phase . We have also shown that catalytically inactive Trt1-D743A shows increased and constitutive binding to telomeres ( Figure 6 ) , consistent with the notion that telomerase is preferentially recruited to short telomeres . The notion that fission yeast utilizes the differential arrival of leading and lagging strand polymerases to control Rad3ATR-dependent Ccq1 Thr93 phosphorylation and Trt1TERT recruitment can explain why mutations in Polε lead to shorter telomeres while mutations in Polα and Polδ lead to longer telomeres [48] . Since mutations in Polε would likely delay leading but not lagging strand synthesis , cells would accumulate less ssDNA at telomeres , and as a result , recruit less Rad3ATR and Trt1TERT . Conversely , mutations in Polα and Polδ would lead to increased ssDNA , and more robust recruitment of Rad3ATR and telomerase . Effects on differential strand synthesis at telomeres could also explain why rif1Δ rap1Δ cells have longer telomeres than rap1Δ cells [8] , since the loss of Rif1 is expected to advance the arrival of Polε [42] , further expanding the differential strand synthesis over rap1Δ cells . Differences in Polα binding ( Figure 2C ) could also explain why rap1Δ cells retain S phase-specific G-tail elongation while taz1Δ cells show elongated G-tails throughout the cell cycle [34] . Even though budding yeast cells have significantly diverged in telomere protein composition from fission yeast or mammalian cells [4] , mutations in Polε also cause telomere shortening while mutations in Polα cause telomere lengthening in budding yeast [49] , [50] . Thus , differential regulation of leading and lagging strand synthesis could have evolutionarily conserved roles in telomerase regulation . Studies in mammalian cells have also found that lagging strand synthesis is significantly delayed [51] and regulated by CST [20] , [21] . Thus , we believe that our current findings are also relevant in understanding how shelterin and CST regulate telomere maintenance in mammalian cells .
Fission yeast strains used in this study were constructed by standard techniques [52] , and they are listed in Table S2 . For taz1Δ::ura4+ , taz1Δ::LEU2 , rap1Δ::ura4+ , poz1Δ::natMX6 and trt1Δ::his3+ , original deletion strains were described previously [8] , [30] , [36] , [53] , [54] . For rad3-kdΔ::kanMX4 , ura4+ marker was swapped with kanMX4 by ( 1 ) PCR amplifying a kanMX4 module from a pFA6a-kanMX4 plasmid [55] using DNA primers UraKan-T1 and UraKan-B1 ( Table S3 ) , and ( 2 ) transforming rad3-kdΔ::ura4+ strain [56] , [57] with the PCR product . For rap1-myc , trt1-myc , pol1-FLAG , pol2-FLAG , myc-rad3 , myc-rad26 , myc-tel1 , rad11-FLAG , ccq1-myc , ccq1-FLAG , tpz1-myc , poz1-myc and stn1-myc , original tagged strains were described previously [9] , [12] , [25] , [57]–[59] . A modified fission yeast strain with leu1-32::[hENT1 leu1+] and his3-D1 his7-366::[hsv-tk his7+] that can be used to efficiently incorporate BrdU has been described previously [60] . Strains that carry trt1-D743A::LEU2 allele at endogenous locus were previously described [61] . A heterozygous diploid strain carrying unmarked trt1-D743A or trt1+ was transformed with a PCR product ( amplified from a trt1-G8-13myc::kanMX6 strain using DNA primers trt1-B29 and trt1-T30 ) to generate cells carrying trt1-D743A-myc . Yeast two/three hybrid assays were performed by mating Saccharomyces cerevisiae MATa ( Y2HGold: MATa trp1-901 leu2-3 , -112 ura3-52 his3-200 LYS2::GAL1 ( UAS ) -GAL1 ( TATA ) -HIS3 GAL2 ( UAS ) -GAL2 ( TATA ) -ADE2 gal4Δ gal80Δ URA3::MEL1 ( UAS ) -MEL1 ( TATA ) -AUR1-C MEL1 ) strains harboring GAL4-DBD ( DNA-binding domain ) plasmids with MATα ( Y187: MATα trp1-901 leu2-3 , -112 ura3-52 his3-200 ade2-101 gal4Δ gal80Δ met- URA3::GAL1 ( UAS ) -GAL1 ( TATA ) -LacZ MEL1 ) strains harboring GAL4-AD ( activation domain ) plasmids , as described in the MATCHMAKER system manual ( Clontech ) . Plasmids used in yeast two/three hybrid assays are listed in Table S4 . Pulsed-field gel electrophoresis to analyze telomere fusions in G1 was performed as previously described [31] , [62] . Telomere probe used in Southern blot and dot blot-based ChIP was generated as previously described [54] , and rDNA probe used to determine telomere length correction factor for dot blot-based ChIP analysis was generated using PCR with primers listed in Table S3 [63] , [64] . Primers used in real-time PCR-based ChIP assays are also listed in Table S3 . ChIP samples were analyzed with quantitative real-time PCR or dot blot with telomeric probe as previously described [10] , [25] . BrdU incorporation was monitored as previously described [25] , [65] . Error bars in all plots represent standard error of the mean ( SEM ) from multiple independent experiments . For western blot and ChIP assays , either monoclonal anti-myc ( 9B11 , Cell Signaling ) or monoclonal anti-FLAG ( M2-F1804 , Sigma ) antibodies were used . Anti-Cdc2 antibody ( y100 . 4 , Abcam ) was used in western blot analysis as a loading control . Ccq1 Thr93 phosphorylation was monitored using phospho- ( Ser/Thr ) ATM/ATR substrate antibody ( 2851 , Cell Signaling ) as previously described [10] . While not specifically raised against a Ccq1 Thr93 phosphopeptide , our previous analysis indicated that the phospho- ( Ser/Thr ) ATM/ATR substrate antibody can specifically detect a Ccq1 Thr93 phosphopeptide , and detect a band corresponding to immunoprecipitated Ccq1 that is eliminated in ccq1-T93A mutant in western blot analysis [10] . Thus , although we cannot completely eliminate the possibility that this antibody recognizes phosphorylation on other site ( s ) that might be affected by ccq1-T93A mutation , for sake of simplicity , we denote the signal detected by this antibody as Ccq1 Thr93 phosphorylation in the text . Correction factors for telomere length were established by measuring the hybridization signal intensity of telomere versus rDNA repeats ( telomere/rDNA ) for poz1Δ , rap1Δ and taz1Δ cells compared to wt cells ( Figure S2A and Table S1 ) , using NaOH denatured genomic DNA samples spotted on Nylon membrane by dot blot apparatus . “Telomere length corrected” ChIP values were then calculated by multiplying the background subtracted % precipitated DNA values ( raw % precipitated DNA – no tag control % precipitated DNA ) with the correction factors , and normalizing them to wt . For telomere length corrected cell cycle ChIP , values were normalized to the peak binding level of wt cells in late S/G2-phase . While it may not be a perfect solution , the use of correction factors provided better estimates of changes in protein binding to chromosome ends for cells carrying significantly longer telomeres than wt cells . For telomere length corrected ChIP data , SEM of telomere length corrected ChIP ( SEMQ ) was calculated as ( A = background subtracted ChIP; SEMA = SEM of background subtracted ChIP; B = telomere correction factor; SEMB = SEM of telomere correction factor ) . In order to determine the statistical significance of our data , two-tailed Student's t-tests were performed , and p-values≤0 . 05 ( ≥95% confidence level ) were considered as statistically significant differences . | Stable maintenance of telomeres is critical to maintain a stable genome and to prevent accumulation of undesired mutations that may lead to formation of tumors . Telomere dysfunction can also lead to premature aging due to depletion of the stem cell population , highlighting the importance of understanding the regulatory mechanisms that ensure stable telomere maintenance . Based on careful analysis of cell cycle-regulated changes in telomere association of telomerase , DNA polymerases , Replication Protein A , checkpoint kinases , telomere protection complex shelterin , and Stn1-Ten1 complex , we will provide here a new and dynamic model of telomere length regulation in fission yeast , which suggests that shelterin-dependent regulation of differential arrival of leading and lagging strand DNA polymerase at telomeres is responsible for modulating Rad3ATR checkpoint kinase accumulation and Rad3ATR-dependent phosphorylation of shelterin subunit Ccq1 to control telomerase recruitment to telomeres . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Fission Yeast Shelterin Regulates DNA Polymerases and Rad3ATR Kinase to Limit Telomere Extension |
There is increasing consensus that males are more vulnerable than females to infection by several pathogens . However , the underlying mechanism needs further investigation . Here , it was showed that knockdown of androgen receptor ( AR ) expression or pre-treatment with 5α-dihydrotestosterone , the AR agonist , led to a considerably dysregulated Kaposi’s sarcoma-associated herpesvirus ( KSHV ) infection . In endothelial cells , membrane-localized AR promoted the endocytosis and nuclear trafficking of KSHV . The AR interacted with ephrin receptor A2 ( EphA2 ) and increased its phosphorylation at residue Ser897 , which was specifically upregulated upon KSHV infection . This phosphorylation resulted from the AR-mediated recruitment of Src , which resulted in the activation of p90 ribosomal S6 kinase 1 ( RSK1 ) , which directly phosphorylates EphA2 at Ser897 . Finally , the EphA2-mediated entry of KSHV was abolished in a Ser897Asn EphA2 mutant . Taken together , membrane-localized AR was identified as a KSHV entry factor that cooperatively activates Src/RSK1/EphA2 signaling , which subsequently promotes KSHV infection of both endothelial and epithelial cells .
Males of many species are more susceptible than females to infections caused by parasites , fungi , bacteria , and viruses . Among humans , there is a reported male predominance in the prevalence and lethality of infections with various pathogens . This may reflect different exposures and immune responses , or even differences in genetic susceptibility between genders [1–3] . Sex-based differences typically become apparent after puberty , which suggests a role of steroid hormones in pathogenesis . Most current studies have investigated this discrepancy in terms of gender-specific immune responses , and the results showed that females have a greater ability to produce immune responses against infections . 17β-Estradiol regulates the activity of immune cells , including lymphocytes , macrophages , granulocytes , and mast cells [4 , 5] . A lack of the inhibitory factor CD200R in females leads to Toll-like receptor 7-mediated activation of interferon-α , which accounts for higher immune status in females , at least in a murine model [6–8] . Additionally , sex hormones can directly affect pathogen infections . Higher serum androgen levels and an androgen receptor ( AR ) gene containing shorter CAG repeats ( which lead to higher AR activity ) have been clinically linked to higher risks of hepatitis B virus ( HBV ) -mediated hepatocellular carcinoma ( HCC ) [9] . The AR increases HBV genome replication by binding to two androgen-responsive elements that are located in enhancers I and II of HBV , which strongly implicates male gender as a risk factor for HCC development [10 , 11] . Correspondingly , estrogen and the estrogen receptor repress the transcription of HBV genes by binding competitively with hepatic nuclear factor 4α to enhancer I [12] . However , whether male sex hormones function in the pathogenesis of other human viruses remains largely unknown . Kaposi’s sarcoma ( KS ) , at least the classical and endemic types , occurs disproportionately in men [13–18] . The age-standardized incidence rate of KS was 12 . 3 and 4 . 6 per 100 , 000 in African males and females , respectively . In older age groups , KS was about 10 times more common in males [19] . Regarding the gender-associated seroprevalence of KS-associated herpesvirus ( KSHV ) , the causative agent of KS , a recent evidence-based meta-analysis indicated that KSHV preferentially infects males in Africa [20] , and a significantly higher quantity of KSHV DNA has been detected in men than women [21 , 22] . These data strongly suggest that male hormones may play a role in KSHV infection and pathogenesis . However , whether and how the hormone system is involved in these processes still remains unknown . The classical role of the AR is that of a steroid hormone-activated transcription factor . Intracellular AR translocates into the nucleus and then stimulates the transcription of androgen responsive genes after binding its hormone ligand . However , another category of membrane-localized AR in Lipid Rafts ( LRs ) was identified a decade ago , and its biological significance remains unknown [23 , 24] . Here , we demonstrate that membrane-localized AR can promote KSHV infectivity , especially at the early entry stage . Both AR and 5α-dihydrotestosterone ( DHT ) , the agonist of AR , promoted KSHV infection , as determined by a quantitative real-time polymerase chain reaction ( qRT-PCR ) assessment of the copy number of the KSHV genome and its transcripts . This effect was mediated by association with , and the consequent phosphorylation of ephrin receptor A2 ( EphA2 ) , one of the major KSHV entry receptors [25 , 26] . The specific residue Ser897 of EphA2 was identified as an essential phosphorylation site responsible for KSHV entry . Interestingly , the Ser897 phosphorylation of EphA2 can be induced by the AR-mediated recruitment of Src , which led to the activation of the kinase p90 ribosomal S6 kinase 1 ( RSK1 ) , which directly phosphorylates EphA2 . These findings demonstrated that male sex hormones facilitate KSHV primary infection through a Src/RSK1/EphA2 Ser897 signaling cascade and may imply a novel mechanism for gender disparity in KS incidence .
As the common function of LRs in promoting KSHV primary infection [27] , we speculated that co-localized AR may play a concordant role in KSHV infection of target cells . KSHV had a broad tropism in vivo in a variety of cell types such as endothelial cells , epithelial cells , monocytes and keratinocytes . Herein , primary human umbilical vein endothelial cells ( HUVECs ) and a culture of epithelial-cell origin ( SLK cells ) were employed to analyze the role of the AR in KSHV infections through RNA interference; a small interfering RNA ( siRNA ) targeting EphA2 was used as a positive control since EphA2 is known to be the entry receptor of KSHV infection for these cells [25 , 26] . The inhibitory effect of the AR siRNA was demonstrated by reduced AR expression ( Fig 1a–1c ) and the consequent inability to upregulate the transcription of AR target genes , the prostate-specific antigen ( PSA ) and nuclear receptor coactivator 2 ( NCOA2 ) genes [28] ( Fig 1b and 1c ) . The specificity of AR detection was confirmed by its abundance of 110 kDa full-length isoform and the typical nuclear localization in androgen-sensitive cells ( S1 Fig ) . Furthermore , AR inhibition led to a dramatic reduction of KSHV infection , as determined by measuring the number of internalized KSHV copies of the LANA gene in HUVECs and SLK cells ( Fig 1d ) . Only 398 , 000 viral copies were internalized in AR siRNA-treated HUVECs , compared with 555 , 000 KSHV copies in control siRNA-treated HUVECs ( a 28% reduction ) , while those for SLK cells were 194 , 000 and 451 , 000 , respectively ( a 57% reduction ) ( Fig 1d ) . Expectedly , the mRNA levels of viral genes in HUVECs with AR knockdown was considerably decreased , as indicated by 54 , 44 , and 46% reductions in the transcription of the LANA , replication and transcription activator ( RTA ) , and polyadenylated nuclear RNA ( PAN ) genes , respectively ( Fig 1e ) , and by 85 , 71 , and 60% reductions , respectively , in SLK cells ( Fig 1f ) . Accordingly , compared with the control groups , at 48 h post-infection ( p . i . ) , we observed a dramatic inhibition of nuclear LANA immunostaining in AR and EphA2 siRNA-treated SLK cells ( Fig 1g and 1h ) . Importantly , we found that not only the AR , but also its ligand , DHT , were capable of increasing KSHV infection in HUVECs and SLK cells ( Fig 2 ) . This was also validated in lymphatic endothelial cells ( LECs ) , another well-established endothelial cell model for in vitro KSHV infection ( S2 Fig ) . Collectively , these results strongly suggest that both the AR and its ligand are able to facilitate KSHV infection in various cell types . KSHV infection of endothelial cells consists of multiple steps [29 , 30] , therefore it is necessary to define the stage at which AR facilitated KSHV infection . As lipid rafts ( LRs ) , where AR is located , have been shown to be essential for the post-binding and entry stages of KSHV infection [27 , 31] , we hypothesized that the AR may also contribute to this process . It was reported that KSHV enters the cells through endocytosis and it should retain its envelope immediately after internalization but lose it when subsequent fusion with endosomal membrane occurs , thus the glycoprotein B ( gB ) is one of the viral markers to indicate the early stage of KSHV endosome trafficking [32] . Here , the intracellular gB staining was used to represent early stage of KSHV entry and trafficking . As shown in Fig 3a , the localization of membrane-localized AR in LRs was confirmed in HUVECs . We further observed that the translocation of the AR from the membrane and cytoplasm into the nucleus occurred immediately upon KSHV infection , as early as 30 min p . i . ( Fig 3a ) . In Fig 3b , the successful internalization and perinuclear accumulation of gB-positive KSHV particles were observed only in permeabilized cells , accompanied by AR nuclear translocation . The specificity of fluorescent gB and AR expression at KSHV early entry stage was verified by involving mock staining for the two molecules , which precisely exclude the contaminant green or red fluorescent signals from rKSHV . 219 virus ( S3 Fig ) . The specificity of fluorescent labeling of LRs was verified by concordant pattern of LR localization between B cells [33] ( S4a Fig ) . And the co-localization between gB and early endosome marker EEA1 ( Early Endosome Antigen 1 ) was identified at 20’ p . i . in HUVEC cells indicating the successful KSHV early endocytosis ( S4b Fig ) . We next examined the efficiency of KSHV endocytosis upon AR siRNA treatment . The results demonstrated a dramatic reduction of the internalized perinuclear staining of gB ( green ) in AR siRNA-treated cells ( Fig 3c ) . As a positive control , treatment with EphA2 siRNA had a greater effect on KSHV internalization and accumulation ( Fig 3c ) . On the contrary , DHT treatment increased the number of KSHV virions that reached the nuclear periphery region in HUVECs ( S4e Fig ) and LECs ( S4f Fig ) as well . Next , we assessed the role of the AR in KSHV binding and entry . Compared with the control siRNA , the AR siRNA inhibited KSHV entry , as determined by significant reductions in the number of internalized KSHV DNA copies in HUVECs ( by 19 . 2% ) and SLK cells ( by 36% ) , but it did not affect KSHV binding ( Fig 3d and 3e ) . In contrast to that , being control virus of Herpes Simplex Virus 1 ( HSV1 ) which independent of EphA2 as cellular receptor [34] , inhibition of either AR or EphA2 had no effect on virus binding and entry ( S4c and S4d Fig ) . Collectively , these results demonstrated that membrane-localized AR can facilitate KSHV infection in the early entry stage , rather than the binding stage . As shown above , the AR participated in KSHV endocytosis . Being a member of the largest family of tyrosine kinase receptors , EphA2 has been defined as a KSHV receptor that is required for virus entry , at least in adherent cells [25 , 26] . It was previously demonstrated that the intracellular kinase domain of the EphA2 receptor is indispensable for KSHV entry [25] , thus we hypothesized the role of AR in mediating the catalysis of EphA2 . Because the specific phosphorylation sites that account for EphA2 phosphorylation have not been reported , we first attempted to identify the sites activated by KSHV infection . The results showed that the phosphorylation of EphA2 at Ser897 , but not that of other tyrosine phosphorylation sites , e . g . , Y594 or Y596/602 , is specifically upregulated by KSHV infection in both HUVECs and SLK cells ( Fig 4 ) . EphA2 was rapidly phosphorylated at Ser897 at 10 min p . i . , and the phosphorylation significantly increased and persisted for 30 min p . i . in HUVEC , whereas a reduction by 90 min p . i . was observed in SLK cells ( Fig 4a and 4b ) . In addition , the AR siRNA had a suppressive effect on EphA2 Ser897 phosphorylation ( Fig 4a and 4b ) . Consistent with this , DHT treatment led to a significant enhancement of the phosphorylation of EphA2 at Ser897 upon KSHV infection ( Fig 4c and 4d ) . Additionally , DHT alone induced EphA2 Ser897 phosphorylation in SLK cells in the absence of KSHV infection ( S5c Fig ) . At 30 and 90 min p . i . , DHT increased the level of phosphorylated EphA2 Ser897 in a dose-dependent manner ( Fig 4d ) . Additionally , DHT treatment further promoted the nuclear translocation of EphA2 that was phosphorylated at Ser897 ( S5a Fig ) , and the effect was synergistically promoted by KSHV infection ( S5b Fig ) . To define the role of EphA2 Ser897 phosphorylation in KSHV entry , we constructed a mutant , named EphA2 Ser897Asn , in which Ser897 was mutated to Asn . This mutation completely abolished the phosphorylation of Ser897 of wild type EphA2 , without affecting the total level of EphA2 ( Fig 5a ) . In addition , the capability of ectopic AR to increase the level of phosphorylated EphA2 was nearly eliminated by the mutant ( Fig 5a ) . Upon KSHV infection , we observed a large amount of KSHV virions in cells that were transfected with a plasmid expressing wild-type ( WT ) EphA2; however , this effect was eliminated in cells that were transfected with a construct expressing the EphA2 Asn897 mutant ( Fig 5b and 5c ) . The results were verified by quantitative analysis to the internalized viral particles represented by red signals ( Fig 5d and S6a Fig ) . Finally , we demonstrated that ectopic AR-induced internalization of KSHV virions is dramatically inhibited by the Ser897Asn mutant , and the results were verified by quantitative analysis showing decreased viral gene expression and KSHV copies ( Fig 5c , 5e and 5f ) . Taken together , these data demonstrate that phosphorylation at Ser897 of EphA2 has a primary role in KSHV entry , and the consistent modulation of Ser897 phosphorylation by the AR and its ligand suggest that it is one of the possible mechanisms by which male hormones facilitate KSHV infection . To explore the mechanism by which AR activates EphA2 , we first attempted to determine whether AR functions by directly binding to EphA2 . As shown in Fig 6a , weconfirmed that AR can co-immunoprecipitates with activated EphA2 during KSHV primary infection , along with Src . The co-localization of the AR with Src was also verified [35] ( S6b Fig ) . Importantly , we observed that the AR associated with EphA2 that is phosphorylated at Ser897 in KSHV-infected SLK cells , which was maximized at 90 min p . i . ( Fig 6a ) . In addition , ectopic EphA2 was efficiently immunoprecipitated by an α-FLAG antibody , which recognizes FLAG-tagged AR , when these proteins were co-expressed in human embryonic kidney 293T cells and the interaction was remained between AR and EphA2 Ser897Asn mutant ( Fig 6b and 6c ) . Moreover , confocal microscopy revealed that the AR co-localized with EphA2 on the cell membrane of HUVECs , as well as in the cytoplasm ( Fig 6d ) . To map the exact EphA2 domains that are responsible for these associations , three glutathione S-transferase ( GST ) fused truncations of EphA2 were accordingly constructed [36] ( Fig 6e ) . Finally , in vitro translated AR specifically bound to the kinase domain of EphA2 , as determined by a GST pulldown assay ( Fig 6f ) . Taken together , these studies suggest that the AR may promote KSHV endocytosis as a host entry factor by interacting with EphA2 and host signaling molecules . Next , we explored the molecular mechanism by which AR activates EphA2 . We speculated that RSK1 may be involved in this process because it phosphorylates EphA2 at Ser897 [37 , 38] ( Fig 7a ) . It was demonstrated that RSK1 is a critical downstream signaling component of the AR , as indicated by the near elimination of RSK1 activation by the AR siRNA , compared with control treatments ( Fig 7b ) . We further confirmed that AR forms complex with RSK1 in ectopic expressed 293T cells by co-IP assay ( Fig 7c ) . This regulation leads to the dramatically increased expression of the pEphA2 Ser897 resulting from co-transfection of the recombinant RSK1 plasmid with ectopic AR ( Fig 7d ) . As Src acts as an upstream signal to directly phosphorylate RSKs [37–39] , we hypothesized that the AR may activate RSK1 via Src . We verified that treatment with PP1 , a Src inhibitor , reduced the level of phosphorylated RSK1 and , consequently , downregulated the phosphorylation of pEphA2 Ser897 , without affecting the total level of phosphorylation ( Fig 7e ) . The effectiveness of PP1 was validated by its ability to completely inhibit Src phosphorylation ( S6c Fig ) . The diminishment of the phosphorylation of pEphA2 Ser897 by RSK1 siRNAs further indicated the requirement for RSK1 in Src-mediated activation of EphA2 ( Fig 7f ) . The results further showed that the remarkable enhancement of EphA2 Ser897 phosphorylation , which was induced by ectopic Src and RSK1 , was further promoted by co-transfection with an AR-expressing plasmid ( Fig 7g ) . More importantly , all of these regulatory events were recapitulated during a KSHV infection ( Fig 8a and 8b ) . Compared with ethanol treatment , the dramatic inhibition of EphA2 Ser897 phosphorylation by PP1 was significantly rescued by DHT treatment , both in HUVECs ( Fig 8a ) and SLK cells ( Fig 8b ) . More specifically , EphA2 Ser897 phosphorylation was moderately restored by DHT in SLK cells , except at 90 min p . i . ( Fig 8b ) , while much stronger restoration occurred in HUVECs . The densitometry values that reflect the level of EphA2 phosphorylation at Ser897 ( Fig 8b ) are provided in Fig 8c . Intriguingly , these regulatory events of AR-induced signal pathways were observed in the LR fraction by membrane raft isolation in HUVEC cells ( Fig 8d ) . Upon KSHV infection , the greatly increased expression of AR , along with Src , was specifically identified in cell membrane compartments of HUVEC cells , while extensive upregulation of pEphA2 Ser897 and pRSK1 occurred throughout whole cell lysate ( Fig 8d ) . In summary , we propose that the membrane-localized AR is the major component that mediates the phosphorylation of KSHV EphA2 by associating with Src , and that the Src-recruited RSK1 kinase phosphorylates EphA2 at Ser897 , which is required for successful KSHV entry ( Fig 8e ) .
Sex-based differences result in different immune responses and disease susceptibilities , which lead to a male predominance for certain infectious diseases . However , the mechanisms for this phenomenon remain largely unknown [1–5] . The male predominance of KS in clinical practice has also been well documented [13–18]; however , its mechanism is not well understood either . Herein , to our knowledge , for the first time we demonstrated a mechanism by which male hormones act as a host factor to facilitate KSHV entry by mediating Src/RSK1/EphA2 Ser897 signaling cascades , which implies a novel mechanism for gender disparity in KS . KSHV infection is essential for the development of KS . In the present study , we demonstrated that male sex steroids facilitated the very early step of KSHV infection . Although the vast majority of KS spindle cells are latently infected with the virus , in a small proportion of infected cells the virus undergoes lytic replication leading to the production of mature virus [14 , 17] . Herein , we postulate a novel role of male hormones as internal stimuli that facilitate this secondary infection which may contribute to the pathogenesis . In contrast to direct binding to HBV genome of AR , and promotion of viral replication [10 , 11] , our findings provide the first evidence that membrane-localized AR is exploited for KSHV entry and endocytosis . Although intracellular AR is conventionally recognized as a transcription factor , in the last decade , studies have shown that the actions of androgens are initiated through the stimulation of membrane androgen-binding sites or receptors ( mARs ) [23 , 24 , 39] . Although the molecular identity of these mARs remains elusive , their activation triggers multiple non-genomic signaling cascades , and it regulates numerous cell responses [39] . Here , for the first time , we demonstrated that this counterpart of the AR could be involved in infectious disease . Cell entry by KSHV is a multistep process involving viral envelope glycoproteins as well as several cellular attachment and entry factors [29 , 30] . One of these is EphA2 , which is localized to cell membrane subdomains/LRs , and , therefore , it has great potential for crosstalk with membrane-localized AR that is distributed in these specific subcompartments [25 , 26] . A lack of the intracellular kinase domain of EphA2 leads to a dramatic ( greater than 70% ) decrease in KSHV infection rates [25] . To our knowledge , the present study is the first to demonstrate that EphA2 phosphorylation at Ser897 is primarily responsible for this effect . Additionally , we identified a novel mechanism of the AR/Src/RSK1 signaling cascade , accounting for EphA2 phosphorylation . Gao lab and Chandran lab demonstrated that vital host pathways ERK ( Extracellular signal-Regulated Kinase ) /MAPK ( Mitogen-Activated Protein Kinase ) and FAK ( Focal Adhesion Kinase ) /PI3K ( Phosphoinositide 3-kinase Phosphatidylinositol-4 , 5-bisphosphate 3-kinase ) /PKC ( Protein kinase C ) are essentially required for KSHV primary infection [40–42] . RSKs are downstream effectors of the Ras-ERK/MAPK signalling cascade [38] , thus these pathways could also be alternative mechanisms for AR-mediated promotion of KSHV infection . As EphA2 had been identified as KSHV receptor for endothelial cells [25] , it turned out to be the major candidate by male sex steroids in our first attempt . KSHV utilizes heparan sulfate , integrins , xCT ( Cystine Transporter ) and DC-SIGN ( Dendritic Cell Specific Intracellular adhesion molecule-3 ( ICAM-3 ) Grabbing Non- integrin ) in context of target cell types [43] , therefore they could also be hijacked by AR and would be interesting for future study . Upon KSHV infection , it is notable that besides in LRs fragments , an immediate early accumulation of cytosolic pEphA2 Ser897 and pRSK1 are observed and it may be due to their roles in versatile biological processes other than receptor activation . EphA2 and FAK/Src/PI3K/RhoGTPase pathogenically function in cell cytoskeleton remodeling by providing the mechanical force necessary to complete endocytosis [26 , 31] , suggesting that cytoplasmic RSK1 may participate in the process as well . The phosphorylation of EphA2 at Ser897 is also exploited by Chlamydia trachomatis to activate phosphatidylinositol-4 , 5-bisphosphate 3-kinase ( PI3K ) signaling to induce apoptosis resistance [44] . Although the phosphorylation of EphA2 at Ser897 has been previously reported to function in the ligand-independent promotion of tumor malignant progression [37 , 38 , 45] , its role in infectious diseases need further investigation . Therefore , it may represent a new candidate for drug development for the prevention of KSHV infections , at least in high-risk populations . Male hormones contribute to the male predominance in certain infectious diseases through various mechanisms , either having an indirect function by hijacking immune cells [4–8] , or by physical interaction with pathogens [10–11] . For the first time , this study demonstrated that the male sex hormones acted as host cofactors in the pathogenesis of primary KSHV infection , which implies a novel mechanism for gender disparity in KS . Considering that EphA2 is also the receptor for some other viruses such as hepatitis C virus ( HCV ) and that it is a signaling hub [36 , 46] , our findings may be relevant to other viral diseases and to endocrine-associated oncogenesis .
HUVECs ( ATCC CRL-1730 ) were cultured in complete endothelial basal medium-2 ( Lonza ) . LEC were purchased from PromoCell ( C-12216 ) and cultured with Endothelial Cell Growth Medium MV2 kit ( C-22121 , PromoCell ) . BJAB ( KSHV-negative B cells ) and BCBL1 ( KSHV-positive PEL cells ) were generously provided by Dr . Erle S Robertson ( University of Pennsylvania , USA ) and were maintained in Roswell Park Memorial Institute 1640 medium ( HyClone ) supplemented with 10% FBS ( HyClone ) . KS-derived SLK epithelial cell lines and doxycycline inducible recombinant KSHV . 219 harboring SLK ( iSLK . 219 ) cell lines was established by J . Myoung and D . Ganem , and was kindly provided by Fanxiu Zhu ( Florida State University ) . iSLK . 219 cells were cultured in DMEM supplemented with 10% fetal bovine serum , 1% penicillin-streptomycin , 1 μg/ml puromycin , 250 μg/ml G418 , and 1 mg/ml hygromycin B . Androgen-sensitive human prostate adenocarcinoma cells ( LNCap ) ( TCHu173 ) , androgen-independent prostate cancer cells ( PC3 ) ( TCHu158 ) and 293T cells ( GNHu17 ) were purchased from cell bank/stem cell bank of Shanghai Institutes of Biological Sciences , Chinese Academy of Sciences ( Shanghai , China ) . SLK , LNCap , PC3 and 293T cells were maintained in Dulbecco’s modified Eagle’s medium ( DMEM ) ( HyClone ) supplemented with 10% fetal bovine serum ( FBS ) ( HyClone ) . Before DHT treatment , charcoal-stripped FBS ( 10%; CD-FBS ) ( Sigma–Aldrich , St . Louis , MO , USA ) in basic medium was pre-utilized for cell culture for 24 h , from which endogenous hormones and growth factors had been removed . The antibodies and reagents were as follows: anti-AR antibody ( ab74272 , Abcam ) , anti-EphA2 antibody ( ab54968 , Abcam ) , anti-phospho-EphA2 ( Y594 ) ( 3970S , Cell Signaling ) , anti-phospho-EphA2 ( Y596/602 ) ( 92590 , Millipore ) , anti-phospho-EphA2 ( S897 ) ( 6347S , Cell Signaling ) , anti-Src pan antibody ( 44656G , Invitrogen ) , anti-phospho-Src antibody ( S418 ) ( 44660G , Invitrogen ) , anti-phospho-RSK1 antibody ( T539+S363 ) ( Cy5344 , Abways ) , anti-RSK1 antibody ( ab32526 , Abcam ) , and anti-KSHV ORF8 antibody ( ab36599 , Abcam ) , anti-LANA monoclonal antibody ( LN53 , ABI ) ; anti-EEA1 antibody ( ab2900 , Abcam ) , anti-LANA antibody ( 1B5 ) was prepared in our laboratory . Secondary antibodies ( Thermo Fisher Scientific ) included goat anti-rabbit antibodies conjugated with Alexa Fluor 488 [A-11094] , 555 [A27017] , and 680[A27020] ) , and goat anti-mouse antibodies conjugated with Alexa Fluor 488 [A-11001] , 555 [A-21422] , and 680 [A-28183] ) . DHT ( D-073 , Sigma–Aldrich ) , PP1 ( sc-203212 , Santa Cruz Biotechnology , Dallas , TX , USA ) , Protease Inhibitor Cocktail Set III ( 539134 , Millipore ) , Phosphatase Inhibitor Cocktail ( sc-45044 , Santa Cruz Biotechnology ) , doxycycline hyclate ( D9891-25G-9 , Sigma–Aldrich ) , hygromycin ( V900372-1G , Sigma–Aldrich ) , puromycin ( OGS541-5UG , Sigma–Aldrich ) , TPA ( P1585 , Sigma-Aldrich ) and G418 disulfate salt ( A1720-5G , Sigma–Aldrich ) ; control siRNA ( fluorescein isothiocyanate conjugate ) -A ( sc-36869 , Santa Cruz Biotechnology ) , AR siRNA ( sc-29204 , Santa Cruz Biotechnology ) , EphA2 siRNA ( sc-29304 , Santa Cruz Biotechnology ) , and RSK1 siRNA ( 6309S , Cell Signaling Technology ) ; Lipofectamine 2000 ( 11668019 , Thermo Fisher Scientific ) , 4' , 6-diamidino-2-phenylindole ( DAPI ) ( Beyotime , c1002 ) ; anti-FLAG M2 affinity gel ( A2220-5 ml , Sigma-Aldrich ) , recombinant protein A/G agarose ( 15948-014/15920-010 , Invitrogen ) , glutathione Sepharose 4B ( 17-0756-01 , GE Healthcare ) ; Vybrant LR Labeling Kits ( Life Technologies , v-34404 ) , Caveolae/Rafts Isolation Kit ( Sigma , CS0750 ) , the TNT T7 Quick Coupled Transcription/Translation System ( L1170 , Promega ) , the Accuprep Genomic DNA Extraction Kit ( k-3032 , Bioneer ) , the Mut Express II Fast Mutagenesis Kit V2 ( C214-01 , Vazyme ) , Amicon Ultra-4 Centrifugal Filter Units ( Millipore , UFC801008 ) and collagen type I cell ware coverslips ( 354089 , BD Biosciences ) . Plasmids: The plasmids pAR-FLAG ( expressing FLAG-tagged AR ) , pEphA2-copGFP ( expressing EphA2 ) , and pRSK1-HA ( expressing hemagglutinin-tagged RSK1 ) comprising amino acids 512–918 ( ref [M23263 . 1] for AR ) or full length sequences ( ref[NM_004431 . 3] for EphA2 and ref[EF043873 . 1] for RSK1 ) were generated by PCR amplification of the respective fragment from cDNAs . The resulting amplicons were inserted into the pCDH-CMV-SF-IRES-Blast , pCDH-CMV-MCS-EF1-copGFP ( System Biosciences , SBI ) , and pCMV-HA vectors ( Clontech ) , respectively . The plasmid pSRC-FLAG ( expressing FLAG-tagged SRC ) comprising full length sequence ( ref [NM-005417] for SRC ) was generated by PCR amplification of a target fragment from SRC expressing bacteria ( X-GWDD70769 , Genechem ) . AR-pcDNA3 . 1 ( + ) -HA was constructed by subcloning the HA-AR fragment into pcDNA3 . 1 ( + ) from pHA-AR which was constructed by subcloning the AR fragment from pAR-FLAG into the pCMV-HA vector . Three truncations of EphA2 were fused to GST in the pGEX-4T-1 backbone vector ( GE Healthcare ) , and they comprised amino acids 1–519 ( the extracellular domain ) , 1–558 ( the extracellular plus transmembrane region ) , and 613–871 ( the kinase domain ) . The EphA2 Ser897Asn mutant was obtained by site-directed mutagenesis of the pEphA2-copGFP plasmid . The reporter plasmid pLANA-pGL2 . 0 was described previously . The plasmid pHSV1-UL30-C comprising part of the C-terminal of HSV1-UL30 ( NC_001806 , 65581–66480 ) was generated by PCR amplification of the according fragment from HSV1 genome and inserted into the pCDH-CMV-SF-IRES-Blast . All of the primers are summarized in S1 Table . Cells were fixed with 4% paraformaldehyde for 30 min at room temperature , permeabilized with 0 . 5% Triton X-100 , and blocked with 20% normal goat serum ( Life Technologies ) , and then they were reacted with the indicated antibodies , followed by fluorescent dye-conjugated secondary antibodies ( 1:1 , 000 dilution ) . The dilution factor for individual primary antibodies was generally 1:200 . Cell nuclei were stained with DAPI LR labeling was performed according to the manufacturer’s recommendation before the fixation . Briefly , live cells were incubated with the fluorescent cholera toxin B subunit ( CT-B ) conjugate ( 1:1 , 000 dilution ) , followed by crosslinking with the anti–CT-B antibody ( 1:200 dilution ) . The procedures were performed at temperatures below 4°C using chilled complete growth medium . Coverslips were mounted with anti-fade mounting medium ( Beyotime ) and photographed using a digital camera and software ( Olympus FV-1200 ) . Recombinant KSHV . 219 ( rKSHV . 219 ) stocks and wild-type virions were acquired by inducing iSLK-BAC16 cells with doxycycline and inducing BCBL1 cells with 12-O-tetradecanoyl phorbol-13-acetate ( TPA ) individually , as described previously [25 , 26] . Briefly , five days later , the supernatant was collected and cleared of cells and debris by centrifugation ( 1500 g for 1 h at 4°C ) and 0 . 45 um syringe filtration . Virus particles were pelleted by ultracentrifugation ( 25 , 000 × g for 2 h at 4°C ) using a SW28Ti rotor . Various amounts of cell-free virus supernatants were diluted and inoculated into 293T cells that were seeded at approximately 5×105 cells/well into six-well plates 24 h prior to infection . Following inoculation , the plates were immediately centrifuged ( 660 g for 2 h at 30°C ) and then placed back into the CO2 incubator . After the centrifugation , the inoculum was removed and replaced with fresh medium . Cells were collected 24 h later and washed once with cold phosphate-buffered saline ( PBS ) . The percentage of GFP-positive cells was determined using a LSRII fluorescence-activated cell sorter ( BD Biosciences ) . Layout and mean fluorescence parameters were analyzed using FlowJo v4 . 5 . 9 software ( FLOWJO , LLC ) . And DNA numbers for wild-type KSHV were determined by LANA amplification in quantitative qRT-PCR analysis . The multiplicity of infection ( MOI ) was expressed as the number of GFP-positive cells and the normalized LANA expression in each well at the time of analysis . For the low production of KSHV in BCBL1 cells , wild-type virions were only used in immunofluorescent detection for LANA expression . Neither GFP nor RFP of rKSHV . 219 can be detected at 10 to 30 minutes p . i . , thus the recombinant virus was used for immunofluorescent analyzing to KSHV entry . During KSHV infections , different amounts of concentrated virus were added to HUVECs , and SLK and 293T cells at MOIs of 10 , 5 , and 1 , respectively . The inoculation were replaced with the corresponding fresh medium , and the cells were cultured for the indicated times . After removing viruses by washing twice with PBS , the cells were prepared under the indicated conditions and subjected to the following conditions . HUVECs and SLK cells were infected with KSHV for 1 h at 4°C for virus binding and at 37°C for virus entry . Cells were washed twice with PBS to remove unbound viruses , and they were subjected to an additional treatment with 0 . 25% trypsin-EDTA for 5 min at 37°C to remove bound , but non-internalized , viruses for the virus entry analysis . KSHV DNA was extracted according to the manufacturer’s instructions ( the Accuprep Genomic DNA Extraction Kit , Bioneer ) . A total of 200 ng of DNA from each sample was used in a real-time DNA PCR using KSHV LANA gene-specific primers . The LANA gene cloned into the pGL2 . 0 vector ( Promega ) was used as the external standard . Known amounts of the LANA plasmid were used in the amplification reactions along with the test samples . Cycle threshold values were used to generate a standard curve and to calculate the relative copy numbers of viral DNA in the samples . The amount of KSHV DNA was normalized to the amount of purified cellular DNA as determined by primers targeting the glyceraldehyde 3-phosphate dehydrogenase gene . Cells were lysed in TRIzol buffer ( Life Technologies ) , and RNA was isolated according to the manufacturer’s instructions . Reverse transcription was performed with a cDNA Reverse Transcription Kit ( Toyobo ) . Real-time reverse transcription-PCR was performed with a SYBR green Master Mix kit ( Toyobo ) . Relative mRNA levels were normalized to the level of actin mRNA and calculated by the ΔΔCT method . The primer sequences are summarized in S1 Table . HUVECs and SLK cells were seeded into six-well plates and transfected at ~80% confluency with siRNA pools from Santa Cruz Biotechnology targeting either the AR or EphA2 . siRNAs were transfected using Lipofectamine 2000 ( Thermo Fisher Scientific ) according to the manufacturer’s instructions . The concentrations for HUVECs and SLK cells were 100 nM and 200 nM , respectively . Cells were cultured at 37°C for 6 h , washed , and maintained for another 18 h . siRNA targeting RSK1 was transfected into 293T cells using Lipofectamine 2000 . The concentration of siRSK1 was 150 nM . Recombinant expression plasmids were transfected into 293T cells using polyethyleneimine for 12 h , and cells were continually cultured in fresh medium for 36 h before collection . Cell lysates were prepared in radioimmunoprecipitation assay ( RIPA ) buffer ( 50 mM Tris-HCl [pH 7 . 4] , 150 mM NaCl , 0 . 5% Triton X-100 ) containing protease and phosphatase inhibitors . Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) and transferred to polyvinylidene difluoride membranes for immunoblotting with the indicated antibodies . Cells were lysed in RIPA buffer containing protease and phosphatase inhibitors for 2 h on ice , with brief vortexing every 10 min . The lysates were centrifuged at 15 , 000 g for 20 min at 4°C to remove cell debris . Supernatants were incubated with the indicated antibodies or affinity beads at 4°C for 2 h , with gentle rotation . The immunoprecipitates were separated by SDS-PAGE and analyzed by immunoblotting . GST fusion proteins were expressed in Escherichia coli BL21 and purified using glutathione-Sepharose 4B ( GE Healthcare ) according to the manufacturer’s instructions . For the pulldown assays , glutathione beads were incubated with purified GST-tagged proteins in RIPA buffer containing 0 . 5% bovine serum albumin at 4°C overnight , with gentle rotation . In vitro-translated AR protein , which was produced by the TNT coupled transcription/translation system ( Promega ) , was further incubated for 2 h . Bound proteins were analyzed by SDS-PAGE and immunoblotted with an anti-HA antibody . 2×107 of HUVEC cells at 80–90% confluence were infected by KSHV at an MOI of 10 for 10' and 30' , or left uninfected , and were washed twice with ice-cold PBS , then subjected to the isolation of the microdomains from the cell plasma membrane according to the manufacturer’s instructions ( CS0750 , Sigma ) . All the work was performed in a cold room . Briefly , a cell lysate was prepared by adding lysis buffer containing Triton X-100 and incubating on ice for 1h . Density gradients at 0% , 20% , 25% , 30% and 35% were prepared using the recommended amounts of the cell lysate , lysis buffer and OptiPrep , and then centrifuged at 200 , 000 g using a SW28Ti rotor ( CP80NX , HITACHI ) for 4h at 4°C . Each fraction was carefully collected from the top to the bottom of the ultracentrifuge tubes . The LR subcompartment ( at 20% and 30% OptiPrep layers ) fractions were condensed using a centrifugal filter ( Millipore Amicon Ultra , UFC801008 ) and were detected by immunoblot assay . Data are expressed as means ± standard errors of the means ( SEM ) . One-way ANOVA analysis , paired and unpaired Student’s t-tests were performed with GraphPad Prism software ( GraphPad Software , Inc . , 7825 Fay Avenue , Suite 230 , La Jolla , CA 92037 USA ) . | Although KS incidence is higher in males , which correlates with higher seroprevalence and viral DNA levels in the blood , little is known whether male sex steroids contribute to this disparity . In the present study , we have confirmed the role of both AR and its ligand in promoting KSHV primary infection in target cells . Specifically , AR inhibition led to a dramatically decreased number of perinuclear-accumulated virus particles during early KSHV entry stage . Mechanically speaking , the effect was resulted from the interaction of AR with known KSHV receptor EphA2 and stimulating signal transduction . The AR recruited Src , activated RSK1 , and then increased EphA2 phosphorylation at residue Ser897 , which is prerequisite for successful KSHV infection . Our study provides for the first time a unique insight into why KSHV may have a higher prevalence in males . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"gene",
"regulation",
"pathogens",
"293t",
"cells",
"biological",
"cultures",
"cell",
"processes",
"microbiology",
"plasmid",
"construction",
"viruses",
"dna",
"vir... | 2017 | Male hormones activate EphA2 to facilitate Kaposi’s sarcoma-associated herpesvirus infection: Implications for gender disparity in Kaposi’s sarcoma |
A non-targeted metabolomics-based approach is presented that enables the study of pathways in response to drug action with the aim of defining the mode of action of trypanocides . Eflornithine , a polyamine pathway inhibitor , and nifurtimox , whose mode of action involves its metabolic activation , are currently used in combination as first line treatment against stage 2 , CNS-involved , human African trypanosomiasis ( HAT ) . Drug action was assessed using an LC-MS based non-targeted metabolomics approach . Eflornithine revealed the expected changes to the polyamine pathway as well as several unexpected changes that point to pathways and metabolites not previously described in bloodstream form trypanosomes , including a lack of arginase activity and N-acetylated ornithine and putrescine . Nifurtimox was shown to be converted to a trinitrile metabolite indicative of metabolic activation , as well as inducing changes in levels of metabolites involved in carbohydrate and nucleotide metabolism . However , eflornithine and nifurtimox failed to synergise anti-trypanosomal activity in vitro , and the metabolomic changes associated with the combination are the sum of those found in each monotherapy with no indication of additional effects . The study reveals how untargeted metabolomics can yield rapid information on drug targets that could be adapted to any pharmacological situation .
Human African trypanosomiasis ( HAT ) is a parasitic infection in sub-Saharan Africa transmitted by tsetse flies . Its causative agent is the flagellated protozoan Trypanosoma brucei , with two sub-species , T . b . gambiense and T . b . rhodesiense responsible for human disease [1] , [2] . There are five drugs in use against HAT . Of these five , only eflornithine has a confirmed mode of action ( MOA ) , namely , inhibition of ornithine decarboxylase ( ODC ) [3] , [4] with concomitant perturbation of the polyamine pathway . In addition to the four licensed drugs , nifurtimox has been recommended by the World Health Organisation for use against late-stage disease in combination with eflornithine [5] . The MOA for nifurtimox has , however , yet to be fully elucidated . For many years it was presumed to exert its action through the generation of oxidative stress associated with reduction of the nitro group with subsequent reduction of oxygen to toxic reactive oxygen species [6] , [7] . In trypanosomes polyamines serve an unusual role in combining with glutathione to create the metabolite trypanothione [8] , which carries out many of the cellular roles usually attributed to glutathione in other cell types , including protection against oxidative stress . This indicated that eflornithine , which inhibits polyamine biosynthesis [9] , [10] and subsequently trypanothione biosynthesis , would synergise with nifurtimox as result of a reduced ability of cells to deal with oxidative stress . However , the data that lead to the conclusion that nifurtimox causes oxidative stress is inconclusive [7] and recent evidence shows that nifurtimox is activated upon metabolism to an open chain nitrile [11] and that this nitrile is as toxic as the parent drug . In mice there was no indication that either drug enhanced uptake of the other into brain [12] , indeed eflornithine diminished brain penetration of nifurtimox in short term uptake assays . Moreover , isobologram analysis indicated that the two drugs were not synergistic in vitro [13] . It is very rare for a new chemotherapeutic agent to be licensed without prior knowledge of its MOA . In 2009 , 19 drugs were approved by the FDA's centre for drug evaluation and research in the US , only one of which had a wholly unknown MOA [14] . A knowledge of the MOA reduces the risk of unexpected toxicity and allows synergism and resistance mechanisms to be predicted . Currently , the MOA of a drug is predicted using expensive and time-consuming enzyme-based assays , followed by targeted analyses of whether cellular death is associated with changes consistent with loss of the predicted target . Metabolomics is a relatively new technology that enables the simultaneous identification of hundreds of metabolites within a given system . In principle , if an enzyme is inhibited by a drug then the concentration of substrate should rise within a system and the concentration of product fall . We have recently introduced metabolomics approaches to investigate metabolism in trypanosomes [15]–[19] . Here we use our metabolomics platform to test the mode of action of eflornithine ( an ornithine decarboxylase suicide inhibitor that has had its MOA validated ) and nifurtimox ( a drug for which the MOA is incompletely understood ) . The combination therapy was also tested to determine any synergy between the drugs at the level of metabolism . Broad changes to cellular metabolomes in response to drug have been determined before [20]–[23] , and an analysis of the effect of eflornithine along with an inhibitor of s-adenosyl methionine decarboxylase were studied using a targeted multi reaction monitoring ( MRM ) approach in Plasmodium parasites , responsible for malaria [24] is one of few studies have focussed on individual changes that can be mapped to specific targets in the metabolic network . Here we reveal that an untargeted LC-MS based metabolomics approach identifies specific changes in the metabolome of trypanosomes that can be related directly to effects induced by these drugs .
Bloodstream form trypanosomes were grown in HMI-9 ( Biosera ) [25] supplemented with 10% foetal calf serum ( Biosera ) , incubated at 37°C , 5% CO2 . Cells for metabolomics assays were grown in 500 cm3 Corning vented culture flasks to a maximum volume of 175 ml per flask . The Alamar blue assay developed by Raz et al . [26] for bloodstream form trypanosomes was used to determine activity of drugs against T . brucei strain 427 . For isobologram analyses , alamar blue assays were conducted for one drug in the presence of three different concentrations ( IC50 , 2× IC50 and 0 . 5× IC50 ) of another drug . A commercial kit ( QuantiChrom , BioAssay Systems ) was used to measure the arginase activity in cell extracts spectrophometrically ( Dynex , wavelength 450 nM ) by the amount of urea produced following manufacturers' specifications . A rapid oil/stop spin protocol , previously described by Carter & Fairlamb [27] , was used to determine uptake of radiolabelled ornithine ( 4 , 5-3H-ornithine , American Radiochemicals ) . Briefly , 100 µl of oil ( 1-Bromodo-decane , density: 1 . 066 g/cm3 ) ( Aldrich ) and 100 µl , 20 µM , 1% ( v/v ) radiolabelled ornithine in CBSS buffer were set up in a tube and 100 µl of cell suspension ( 108 per mL ) for varying lengths of time before stopping the reaction by centrifugation . Alternatively , radiation was used at 1% ( v/v ) and cold ornithine levels were varied , while keeping the incubation time constant at one minute . The resulting cell pellet was flash frozen in liquid nitrogen and the base of the tube containing the pellet was cut into 200 µl of 2% SDS in scintillation vials and left for 30 minutes . Three ml of scintillation fluid was added to each vial and these were left overnight at room temperature . Counts per minute were read on a 1450 microbeta liquid scintillation counter ( Perkin Elmer ) and normalised between samples for the cell density . Michaelis-Menten kinetic analyses were performed using Graphpad Prism 5 software . Metabolite extraction methods were adapted from Leishmania spp extraction techniques developed previously [13] , [28] , [29] . Cultures were kept in log phase growth ( below 1×106/ml ) in the presence of drug . At the time of harvest , 4×107 cells were rapidly cooled to 4°C to quench metabolism by submersion of the flask in a dry ice-ethanol bath , and kept on ice for all subsequent steps . The cold cell culture was centrifuged at 1 , 250 RCF for 10 minutes and the supernatant completely removed . Cell lysis and protein denaturation was achieved by addition of 200 µL of 4°C chloroform∶methanol∶water ( ratio 1∶3∶1 ) plus internal standards ( theophylline , 5-fluorouridine , Cl-phenyl cAMP , N-methyl glucamine , canavanine and piperazine , all at 1 µM ) , followed by vigorous shaking for 1 hour at 4°C . Extract mixtures were centrifuged for two minutes at 16 , 000 RCF , 4°C . The supernatant was collected , frozen and stored at −80°C under argon until further analysis . For heavy metabolite tracking analyses , log phase cells were centrifuged 1 , 250 RCF for 10 minutes and resuspended in CBSS ( 20 mM HEPES , 120 mM NaCl , 5 . 4 mM KCl , 0 . 55 mM CaCl2 , 0 . 4 mM MgSO4 , 5 . 6 mM Na2HPO4 , 11 . 1 mM glucose ) or HMI-9 as outlined in the Results section . Heavy atom labelled amino acids were obtained with 15N incorporation from Cambridge Isotope Laboratories ( L-threonine ( 98% enrichment , one incorporation , alpha-N , cat:NLM-742-0 ) , L-glutamine ( 98% enrichment , one incorporation , alpha-N , cat: NLM-1016-0 ) , L-arginine ( 98% enrichment , four incorporations , allo-N , cat: NLM-396-0 ) , L-ornithine ( 98% enrichment , two incorporations , allo-N , cat: NLM-3610-0 ) , L-lysine ( 95–99% enrichment , one incorporation , alpha-N , cat: NLM-143-0 ) ) or Sigma Aldrich ( L-proline ( 98% enrichment , one incorporation , alpha-N , cat: 608998 ) , L-glutamate ( 98% enrichment , one incorporation , alpha-N , cat: 332143 ) ) . Quenching of metabolism was achieved through rapid cooling and metabolite extraction was conducted as above . Samples were analysed on an Exactive Orbitrap mass spectrometer ( Thermo Fisher ) in both positive and negative modes ( rapid switching ) , coupled to a U3000 RSLC HPLC ( Dionex ) with a ZIC-HILIC column ( Sequant ) as has previously been described [13] . All samples from an individual experiment were analysed in the same analytical batch and the quality of chromatography and signal reproducibility were checked by analysis of quality control samples , internal standards and total ion chromatograms . The few samples that displayed unacceptable analytical variation ( retention time drift ) were removed from further analysis . A standard mix containing approximately 200 metabolites ( including members of the polyamine pathway ) was run at the start of every analysis batch to aid metabolite identification . Untargeted metabolite analysis was conducted with the freely available software packages mzMatch [30] and Ideom ( http://mzmatch . sourceforge . net/ideom . html ) . Raw LCMS data was converted to mzXML format and peak detection was performed with XCMS [31] and saved in peakML format . MzMatch was used for peak filtering ( based on reproducibility , peak shape and an intensity threshold of 3000 ) , gap filling and annotation of related peaks . Ideom was used to remove contaminants and LCMS artefact peaks and to perform metabolite identification . Metabolite identities were confirmed by exact mass ( after correction for loss or gain of a proton in negative mode or positive mode ESI respectively ) and retention time for metabolites where authentic standards were available for analysis , and putative identification of all other metabolites was made on the basis of exact mass and predicted retention time of all metabolites from the KEGG , MetaCyc and Lipidmaps databases [17] . Additional manual curation was performed on all datasets to confirm the identification of metabolites that changed significantly in response to drug treatment , and to remove false-identifications based on the LCMS meta-data recorded in Ideom . In cases where identification was putative , the most likely metabolite was chosen based on available chemical and biological knowledge , however accurate isomer identification is inherently difficult with LCMS data and lists of alternative identifications and meta-data for each identified formula are accessible in the macro-enabled Ideom files ( Figure S1 , Figure S2 , Figure S3 and Figure S4; help documentation available at mzmatch . sourceforge . net/ideom . html ) . Quantification is based on raw peak heights , and expressed relative to the average peak height observed in untreated cells from the same experiment . Unidentified peaks in the LCMS data were also investigated for drug-induced changes , however , after removal of LCMS artefacts and known contaminants , the only reproducible change ( <3-fold ) amongst the unidentified peaks was the appearance of C10H15N3O3S ( mass = 257 . 0834 , RT = 13 . 5 ) in the nifurtimox-treated cells . This mass is in agreement with the saturated open chain nitrile metabolite of nifurtimox .
The IC50 of eflornithine on bloodstream form cells in vitro was 35 µM using a standard alamar blue assay [23] . The IC50 of nifurtimox was 4 µM ( Table 1 ) . The drugs were widely believed to be synergistic given the fact that eflornithine ultimately diminishes polyamine production and in turn production of trypanothione , the trypanosome's principal anti-oxidant , whilst nifurtimox had been proposed to generate oxidative stress [6] , [7] . However , we showed in isobologram analyses that the action of nifurtimox and eflornithine did not synergise when nifurtimox action was assayed in the presence of several fixed concentrations of eflornithine [13] and Fig . 1A . Indeed , an antagonistic effect was seen with a fractional inhibitory concentration of 1 . 61 . To determine the levels at which eflornithine is cytostatic and cytotoxic , time course assays were conducted with drug at various concentrations ( Fig . 1B ) . Eflornithine was found to be cytostatic ( cells remained at the same density even at 500 µM until around 55 hours in drug , when they died ) . There was no overt sign of differentiation to stumpy forms , but as the 427 strain is monomorphic , and thus incapable of the morphological changes associated with differentiation in field isolates , this would not be expected . Nifurtimox , on the other hand , had lysed all trypanosomes by 8 hours in 60 µM drug . It is possible that eflornithine's antagonistic effect could relate to its cytostatic potential , if , for example , nifurtimox activity depends on cellular proliferation . The purine analogues , NA42 and NA134 are also known to be cytostatic [32] and these compounds were tested in combination with nifurtimox and also found to be antagonistic with FICs ( fractional inhibitory concentrations ) of 1 . 40 and 1 . 56 for NA42 and NA134 respectively . DB75 , a known potent trypanocidal agent [33] , on the other hand , was shown to be additive in its activity with nifurtimox ( FIC: 1 . 09 ) . In order to detect molecular targets of eflornithine , a first experiment using sub-lethal levels ( 20 µM ) of drug was used , with the cellular metabolome measured at 0 , 1 , 24 , 48 and 72 hours following exposure to drug . Eflornithine was added to the 427 bloodstream form wild type cell line in the same growth medium in which IC50 values had been determined , so that cells were metabolising as normal apart from the perturbation by the drug . The stringent filtering systems in the mzMatch and IDEOM software reduced the number of peaks in the spectra from several hundred thousand to a few hundred robust signals with putative metabolite identities ( Fig . S1 ) . Most metabolite levels were unaltered over the time points taken , indicating a high level of robustness within the trypanosome metabolome . Ornithine ( mass: 132 . 0899 , RT: 27 . 9 minutes ) , the substrate of eflornithine's known target , ornithine decarboxylase ( ODC ) , was the most significantly modulated metabolite over the time course ( 7 . 5 fold increased at 48 hours ) . Putrescine ( mass: 88 . 1001 , RT: 36 . 91 minutes ) , the product of the ODC reaction was the only known metabolite in the T . brucei metabolite database at KEGG , to significantly decrease ( by 66% at 48 hours ) over time . Acetylated ornithine and putrescine were also detected , and these correlated highly with their non-acetylated counterparts . N-acetyl ornithine ( mass: 174 . 1004 , RT: 15 . 3 minutes ) showed the most striking correlation . N-acetyl-putrescine ( mass: 130 . 1106 , RT: 15 . 5 minutes ) was seen in early samples , but levels rapidly fell below the level of detection ( 1 , 000 ) from an average intensity of 41 , 000 ( peak height ) before drug addition , correlating with the decrease in putrescine . Cells were also treated with 500 µM eflornithine , a lethal dose of the drug . At this dose bloodstream form trypanosomes exhibit division arrest over 48 hours in drug before dying between 48 and 55 hours ( Fig . 1B ) . This was reflected by many more changes to the metabolome ( Figure S2 ) . Changes to polyamine pathway metabolites were again consistent with inhibition of ODC , with significant increases in ornithine and N-acetyl ornithine , and decreases in putrescine and N-acetyl putrescine , observed within 5 hours and maintained for the duration of treatment . Spermidine was significantly decreased by 24 hours , confirming the downstream effect of ODC inhibition on polyamine levels ( Fig . 2 ) . Additional metabolites that significantly increased within 24 hours were putatively identified as N-acetyl spermidine , N-acetyl lysine and N5- ( L-1-Carboxyethyl ) -L-ornithine ( a known bacterial metabolite formed from ornithine and pyruvate , although we are not in a position to rule out its generation as a non-enzymatic liaison between these chemicals during sample preparation ) . These metabolites , along with N-acetyl ornithine , demonstrate metabolic derivitisation of ornithine and other polyamine metabolites , which may be an upregulated process in response to the elevated ornithine levels . Aside from the polyamines , most major decreases in metabolite levels over 24 hours were observed among the phospholipids . Polyamines have previously been shown to be key mediators of membrane stability [34]–[36] , and the lipid degradation observed here is consistent with cell membranes being compromised by polyamine depletion . Furthermore , the majority of metabolites in the cell decrease at the 48 hour time point , indicating a possibility that the cell membrane has been compromised and metabolites may be leaking from the cell during incubation and/or sample preparation . The processing of the cells involves cooling them to 0°C in a dry ice–ethanol bath and two centrifugation steps . These weakened cells are therefore potentially more leaky than cells that have not been compromised by prolonged exposure to eflornithine . Several methionine-related metabolites ( cystathionine , S-adenosyl methionine , methylthioadenosine and methyl-methionine ) increased over the first five hours in drug , which was not reported in previous studies . S-adenosyl methionine is the aminopropyl donor involved in spermidine synthesis , and it is possible that this pathway has been upregulated in response to the declining polyamine levels . Methionine levels do not increase over this time course , however , this may be due to the high concentration of methionine in the growth medium ( 200 µM ) and robust transport [37] masking any changes within the cells . Despite the significant decrease observed for spermidine , levels of trypanothione disulphide were not affected during the first 24 hours of treatment . A significant decrease was observed at 48 hours . The analytical platform used here is not capable of reporting the oxidation state of trypanothione or other thiols . The other significant changes observed during the first 24 hours of eflornithine treatment were not expected . Sedoheptulose ( mass: 210 . 0738 , RT: 14 . 9 minutes ) and sedoheptulose phosphate ( mass: 290 . 0400 , RT: 25 . 4 minutes ) were increased , as well as a metabolite with the chemical formula C7H12O5 ( mass: 176 . 0683 , RT: 7 . 52 minutes ) , putatively identified as propylmalate , but possibly diacetylglycerol or another isomer . Our metabolomics analysis above reveals ODC to be the primary target of eflornithine , as was already clear based on previous work and the design of the compound as a specific inhibitor of the enzyme . Surprisingly , however , we could find no previous work that has focused on the cellular source of ornithine in T . brucei . In many eukaryotes , ornithine is produced from arginine via the enzyme arginase . In Leishmania parasites , which belong to the same taxonomic group as T . brucei , for example , an arginase enzyme has been characterised in some detail [38]–[41] . T . brucei , however , lacks a gene that is syntenic with the known Leishmania arginase . A second gene related to arginase is present in Leishmania and an orthologue is present in T . brucei ( Tb927 . 8 . 2020 ) . This latter predicted enzyme , however , lacks key arginase residues and is currently annotated as a putative agmatinase ( although this also seems unlikely given the lack of conservation of key active site residues ) . We measured arginase activity in Leishmania mexicana extracts and compared this to T . brucei extracts where we show that the trypanosome contains little or no classical arginase activity when compared to Leishmania ( Fig . 3A ) . The absence of a classical arginase raises questions about other potential sources of ornithine in T . brucei . Our experiments did reveal the presence of N-acetyl ornithine in T . brucei , the abundance of which was closely correlated to ornithine . Differences in the retention times between ornithine ( RT = 27 . 9 minutes ) and acetylornithine ( RT = 15 . 3 minutes ) confirm that the two metabolites are not mass spectrometry artefacts . In a variety of bacteria ornithine is produced from glutamate in a pathway that involves N-acetyl ornithine as an intermediate [42] , [43] . We used heavy-nitrogen labelled metabolites to trace whether a similar pathway exists in T . brucei . However , cells incubated with isotopically-labelled extracellular 15N-glutamate failed to accumulate this amino acid to a detectable level . We therefore provided 15N labelled glutamine , which was converted to glutamate ( albeit at a relatively low level of 5% of the non-labelled metabolite ) after two hours and 15N-proline which was converted to glutamate at levels of 3 . 1% of the unlabelled glutamate generated within these cells . However , the heavy atom labelled glutamate was not further converted to ornithine , N-acetyl ornithine or N-acetyl glutamate semialdehyde ( another metabolite of the glutamate to ornithine pathway ) . Furthermore , no orthologues , other than N-acetyl ornithine deacetylase ( Tb927 . 8 . 1910 ) , encoding enzymes of the bacterial pathway could be identified in the trypanosome genome indicating that this pathway is not operative in trypanosomes . Although ornithine is not a component of HMI-9 medium , metabolomics analysis of our medium indicated that the commercial supply we used did contain ornithine and we were able to measure its concentration at 77 µM , using a calibration curve with isotopically labelled ornithine . We therefore measured the ability of 3H-ornithine to enter trypanosomes . This indicated a possible external source of ornithine and we tested the ability of this nutrient to enter trypanosomes . At 10 µM , ornithine was shown to enter bloodstream form T . brucei at a rate of approximately 10 pmol/107 cells/min ( Fig . 3B ) . Kinetic analysis of ornithine transport indicated a Km of 310 µM and Vmax of 15 . 9 pmol/107 cells/min ( Fig . 3C ) . Given that ornithine is present in serum and cerebrospinal fluid ( at concentrations of 54–100 µM in plasma and 5 µM in CSF ( according to the human metabolome database , http://www . hmdb . ca/ ) ) , this would indicate that T . brucei is capable of fulfilling its ornithine requirements directly by transport from the bodily fluids in which it resides . When we used 15N-ornithine externally to trace its metabolism we showed that N-acetyl ornithine , spermidine and trypanothione disulphide when added to cells growing in HMI-9 . 15N-labelled arginine was converted to ornithine when administered in CBSS ( Carter's balanced saline solution ) , but not when administered in HMI-9 growth medium . This suggested that when exogenous ornithine is present , uptake of ornithine is sufficient for polyamine synthesis , but when absent , synthesis from arginine is possible . This was confirmed by the addition of exogenous ornithine in addition to heavy arginine in CBSS , where synthesis of heavy ornithine from arginine was no longer detected ( heavy ornithine being present at 40% of unlabelled ornithine levels when exogenous ornithine was not added under the same conditions ) . The enzymatic route by which arginine is converted to ornithine in the absence of canonical arginase is not known . At the sub-lethal dose of 1 . 5 µM nifurtimox , no significant changes to the metabolome were recorded ( data not shown ) . However , at a lethal dose of 60 µM changes to the metabolome at 0 , 1 , 2 and 5 hours following exposure to drug , were seen ( Figure S3 ) . Nifurtimox ( mass: 287 . 0577 , RT: 5 . 25 minutes ) was observed in all treated samples , in addition to a mass ( mass: 257 . 0834 , RT: 13 . 5 minutes ) consistent with the saturated open chain nitrile metabolite [11] ( Fig . 4A ) which was recently shown to be the end product of the multi-step 2-electron reduction of nifurtimox by type-1 nitroreductase . Previous work in a cell-free system showed the saturated nitrile only after 24 hours of drug exposure to the nitroreductase [11] . Our metabolomics platform allows identification of this metabolite within the cell , and shows the process to be rapid with significant levels detectable at the first , 1 hour time point . The implicit intermediates from this reductive activation cascade , including the unsaturated open chain nitrile proposed to mediate trypanocidal activity , were not observed , indicating either that the reduction is rapid and intermediates in the pathway do not persist at detectable concentrations , or that the reactive intermediates indeed react rapidly with intracellular macromolecules . An exhaustive search of all known metabolites in our database revealed no detectable masses that correspond to a hypothetical adduct between the unsaturated open chain nitrile and any known metabolite . Our metabolomics platform , by definition , was unable to detect the potential formation of adducts between nifurtimox metabolites and macromolecules ( proteins or nucleic acids ) . A number of cellular metabolites were shown to change in abundance over the nifurtimox exposure time course ( Table 2 ) , although 95% of putatively identified metabolites were stable . There was an increase in concentrations of nucleotides and nucleobases ( adenine , deoxyadenosine , AMP , GMP , uracil and UMP ) during the time course , which may result from degradation of RNA and DNA consistent with the hypothesis [11] that the nifurtimox active metabolite ( the open chain nitrile ) binds to macromolecules including nucleic acids , by the ability of the unsaturated nitrile intermediate to act as a Michael acceptor [11] . Glycolysis appeared to be downregulated , with significant decreases in hexose 6-phosphates , and similar trends for glyceraldehyde 3-phosphate and 3-phosphoglycerate . The metabolite that decreased most following nifurtimox treatment was deoxyribose , which may indicate reduced synthesis from the glycolytic intermediates , or could be related to nucleic acid homeostasis . Lipid metabolism was largely unaffected with the exception of decreased levels of monounsaturated ether-linked lysophosphatidylcholines ( 14∶1 , 15∶1 and 16∶1 ) and ethanolamine phosphate . Metabolites of the polyamine pathway were not significantly altered over the nifurtimox time course , although decreased thiol levels ( trypanothione disulphide and glutathionyl-cysteine disulphide ) were observed , suggesting that oxidative stress may be induced on exposure to nifurtimox in agreement with previous studies [6] , [7] , [44] , although the role of this stress in ultimate trypanocidal effect is uncertain . It is noted that this untargeted metabolomics approach is not suited for assessment of redox balance ( as reduced thiols are oxidised during sample preparation and analysis ) , however and it is assumed that the observed disulphide levels are indicative of total thiol levels . The presence of oxidative stress may also explain the observed inhibition of glycolysis [45] , and the decreased levels of arginine phosphate [46] . We also investigated changes to the metabolome associated with exposure to eflornithine and nifurtimox simultaneously ( Figure S4 ) . The metabolome of NECT treated cells was measured using drug levels that were toxic in the monotherapies ( 500 µM for eflornithine and 60 µM for nifurtimox ) and the time points used in the nifurtimox toxicity assay ( 0 , 1 , 2 and 5 hours ) , after which cells died without remaining viable for as long as studied in the eflornithine monotherapy study . The rapid reduction of Nifurtimox ( within 1 hour ) to the saturated open chain nitrile was still observed ( Fig . 4B ) . This indicates that the nitroreductase activity known to be responsible for metabolic activation of nifurtimox [11] is not diminished in a short term response to eflornithine . The combination therapy showed qualitatively most of the same changes that were present in each of the monotherapies alone ( Table 2 and Fig . 5 ) . This indicates that both of the drugs are able to exert their individual effects and no additional effects were apparent using the combination . The eflornithine-induced changes to polyamine pathway metabolites were observed in the combination ( Fig . 2 ) , although the later effects of eflornithine could not be measured as cells died from nifurtimox toxicity before these were apparent . The nifurtimox-induced changes to nucleotides , glycolysis intermediates , deoxyribose and thiols were all observed to a similar extent in the combination treatment .
Understanding how small chemicals interfere with cellular metabolism is a critical part of modern drug development . Here we show how a relatively simple LC-MS based metabolomics platform can be used to identify drug modes of action in the causative agent of human African trypanosomiasis , Trypanosoma brucei . Using each of the drugs currently used in combination as a first line treatment against stage two HAT we reveal how modes of action of drugs can be rapidly ascertained . At low levels of drug ( sub IC50 ) specific changes to the metabolome can be detected as was evidenced with eflornithine . The data reveal very localised changes to the metabolome with little indication of broadly disseminated affects consistent with the theory that metabolic networks are generally robust to perturbations [47] . This study reveals the power of metabolomics for predicting the MOA of compounds with a metabolic ( enzyme inhibition ) mode of action . As ornithine accumulation and putrescine loss were the most significant changes between treated and untreated cells , ornithine decarboxylase emerges as the most likely target for this drug . In this case , the outcome was already known hence the follow up experiments e . g . showing that ODC is essential using gene knockout [48] and that addition of polyamines to the medium can bypass eflornithine toxicity [48] have already been performed . With unknown drugs , of course , these validation experiments are still required once the hypothesis has been set using metabolomics . The presence of the open chain trinitrile in nifurtimox-treated cells confirms the trypanosome-mediated metabolic activation of this drug , as was recently demonstrated following substantial targeted analysis of nifurtimox [11] . It will be of interest to extend studies to other current trypanocides and also to systematically include metabolomics in any test of action for compounds emerging from screens . Eflornithine inhibited ODC relatively quickly with levels of ornithine and putrescine demonstrably altered after just five hours in toxic doses of drug . Trypanothione is a glutathione-spermidine adduct and its overall levels are diminished by around 73% prior to death in the eflornithine study , which is similar to the 66% reduction determined after eflornithine exposure in vivo [10] , but it should be noted that many unrelated metabolites were also diminished at 48 hours . An advantage of the non-targeted metabolomics platform used here over a strictly targeted approach to report on individual metabolites is thus clear . Loss of putrescine and spermidine appears to contribute to cellular toxicity independently of their role in trypanothione biosynthesis as rescue experiments where spermidine is given exogenously to ODC knock down cells were unsuccessful [49] . Our studies indicate that eflornithine is trypanostatic for 48 hours , before killing the parasites after apparently compromising the membrane of the cell , as judged by a general loss of metabolites from the cell and particularly changes in lipid content . Since polyamines have been proposed to help stabilise membrane phospholipids [34] , [35] this could indicate the actual cause of death following reduction in polyamine biosynthesis . In vivo , changes to membrane integrity would also expose new ligands to the immune systems , possibly explaining the need of an active immune system for optimal eflornithine activity [47] . Nifurtimox did not show the same depletion in membrane integrity prior to cell death . The untargeted metabolomics approach was particularly useful for the identification of unexpected metabolites . Acetylated ornithine and putrescine have not been previously described in trypanosomes , and would likely not have been assessed with a classical targeted approach , but these results clearly show the presence of acetylated polyamines , and their dynamic relationships with polyamine levels , with N-acetyl ornithine correlating particularly well with ornithine . This metabolite has an unknown function within trypanosomes but appears to be created directly from ornithine transported into the cell . We have also shown that trypanosomes do not use classical arginase activity comparable to that found in related Leishmania spp . parasites to create ornithine from arginine but do have the ability to transport ornithine which is present in plasma and CSF , indicating that they probably fulfil ornithine needs by acquiring it directly from the host . Interestingly they can , nevertheless , convert arginine to ornithine , but apparently only when exogenous ornithine is not available . An increase in sedoheptulose and sedoheptulose phosphate in eflornithine-treated cells was also of interest . Sedoheptulose phosphate is a seven carbon sugar of the pentose phosphate pathway , formed , along with glyceraldehyde 3-phosphate , from ribose 5-phosphate and xylulose 5-phosphate through transketolase ( Tb927 . 8 . 6170 ) action . Transketolase activity , however , is absent in bloodstream-form trypanosomes [50] , [51] , although it is induced in parasites transforming between bloodstream and procyclic forms . It is possible , therefore , that the increase in sedoheptulose 7-phosphate could relate to transketolase being switched on in relation to the proposed induction of differentiation between slender bloodstream form and stumpy form organisms . Although the s427 strain used here does not differentiate to stumpy forms , other biochemical events such as the induction of enzymes usually repressed in the non-dividing stumpy stage may occur . Nifurtimox treatment did not induce any changes to sedoheptulose or its phosphate's levels . Toxic doses of nifurtimox revealed alterations to levels of various nucleotide , carbohydrate and lipid metabolites . More work is required to ascertain why these metabolites' levels are altered with nifurtimox treatment and how these changes relate to death . However , our data reveal that this metabolomics approach can confirm previous findings that relate oxidative stress to nifurtimox treatment , and demonstrate the production of an open chain trinitrile metabolite in agreement with the proposed mechanism for the drug's selective activity against trypanosomes [11] . We show also that the appearance of this metabolite is relatively fast , being detectable within an hour of exposure . The nifurtimox-eflornithine combination therapy , which was previously assumed to be synergistic , was shown to be mildly antagonistic in vitro . The theory behind synergy was based on the assumption that eflornithine would decease cellular trypanothione levels thus decreasing the ability of these cells to defend against oxidative stress . Since nifurtimox was generally believed to exert an effect through generation of reactive oxygen species [6] , [7] it followed that eflornithine treated cells would show enhanced susceptibility to nifurtimox . However , the metabolic perturbations observed in this study suggest that oxidative stress is not the primary MOA for either drug ( despite some indication of oxidative stress observed with nifurtimox ) , and if nifurtimox actually acts through production of the reactive open chain trinitrile and its ability to covalently modify macromolecules , then the proposed synergy would not exist . It should be noted too , however , that our studies in vitro need not reflect the situation in vivo where pharmacokinetic factors lead to very different exposure of parasites to drug and where other host related factors , not least the immune response , contribute to effects of the drugs , although in mice at least neither drug facilitates entry of the other into the brain . A potential reason why the drug combination is mildly antagonistic in vitro could relate to the activation of nifurtimox and its target based effects depending upon growth status of the cell . There was no evidence that activation of nifurtimox was reduced in the eflornithine co-treated cells . Instead , therefore , it is possible that cells entering a state of reduced growth are less affected by the impact of nifurtimox on energy and nucleic acid metabolism . This hypothesis was supported by the antagonism to nifurtimox seen with the trypanostatic purine analogues NA42 and NA134 [36] . The examples we provide here demonstrate how a relatively simple metabolomics platform can elucidate the mode of action of a drug in a relatively short time frame . This study shows that our metabolomics platform yields hypothesis-free data that confirm the known MOA of eflornithine and create testable hypotheses for the nifurtimox MOA as well as confirming a lack of synergy of NECT . The approach we provide here can be readily adapted for other drugs and cellular systems . | Understanding drug mode of action is of fundamental importance . Of the five drugs in use against human African trypanosomiasis ( HAT ) , convincing evidence on a specific mode of action has been proposed only for the polyamine pathway inhibitor eflornithine . Eflornithine is currently used with nifurtimox as first line treatment of stage 2 CNS-involved HAT . Here , we present a new way of determining the mode of action of a drug by measuring how the levels of small molecules comprising the cellular metabolome are perturbed when exposed to drugs . We show that eflornithine causes the changes in polyamine metabolism previously known to underlie its mode of action . Furthermore , we show that nifurtimox , is rapidly metabolised and significantly alters metabolism . Nifurtimox and eflornithine do not show the predicted synergy with regard to trypanocidal activity and this is reflected in metabolomic analysis where perturbations to the metabolome are additive with no additional changes observed in the combination . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"biology"
] | 2012 | Untargeted Metabolomics Reveals a Lack Of Synergy between Nifurtimox and Eflornithine against Trypanosoma brucei |
Synaptic long-term potentiation ( LTP ) at spinal neurons directly communicating pain-specific inputs from the periphery to the brain has been proposed to serve as a trigger for pain hypersensitivity in pathological states . Previous studies have functionally implicated the NMDA receptor-NO pathway and the downstream second messenger , cGMP , in these processes . Because cGMP can broadly influence diverse ion-channels , kinases , and phosphodiesterases , pre- as well as post-synaptically , the precise identity of cGMP targets mediating spinal LTP , their mechanisms of action , and their locus in the spinal circuitry are still unclear . Here , we found that Protein Kinase G1 ( PKG-I ) localized presynaptically in nociceptor terminals plays an essential role in the expression of spinal LTP . Using the Cre-lox P system , we generated nociceptor-specific knockout mice lacking PKG-I specifically in presynaptic terminals of nociceptors in the spinal cord , but not in post-synaptic neurons or elsewhere ( SNS-PKG-I−/− mice ) . Patch clamp recordings showed that activity-induced LTP at identified synapses between nociceptors and spinal neurons projecting to the periaqueductal grey ( PAG ) was completely abolished in SNS-PKG-I−/− mice , although basal synaptic transmission was not affected . Analyses of synaptic failure rates and paired-pulse ratios indicated a role for presynaptic PKG-I in regulating the probability of neurotransmitter release . Inositol 1 , 4 , 5-triphosphate receptor 1 and myosin light chain kinase were recruited as key phosphorylation targets of presynaptic PKG-I in nociceptive neurons . Finally , behavioural analyses in vivo showed marked defects in SNS-PKG-I−/− mice in several models of activity-induced nociceptive hypersensitivity , and pharmacological studies identified a clear contribution of PKG-I expressed in spinal terminals of nociceptors . Our results thus indicate that presynaptic mechanisms involving an increase in release probability from nociceptors are operational in the expression of synaptic LTP on spinal-PAG projection neurons and that PKG-I localized in presynaptic nociceptor terminals plays an essential role in this process to regulate pain sensitivity .
Plasticity in peripheral nociceptors and their synapses with spinal neurons has been proposed as a cellular basis for the development and maintenance of pain hypersensitivity following peripheral inflammation or nerve injury [1]–[3] . Activation of nociceptive nerve afferents at frequencies relevant to pathological pain states can trigger long-term potentiation ( LTP ) at spinal synapses between nociceptor terminals and spinal neurons projecting nociceptive information to the brain [4] , [5] . Importantly , this form of synaptic plasticity can be evoked by asynchronous activation of nociceptors in vivo [5] , occurs in humans [6] , and is functionally associated with a sensation of exaggerated pain [5] , [6] . Although there is evidence for a requirement of post-synaptic calcium-dependent mechanisms in the induction of LTP at this synapse [5] , the precise mechanisms underlying the expression of spinal LTP are not entirely clear [7] . Synaptic LTP evoked by natural , asynchronous low-rate discharges in C-nociceptors on spino-PAG neurons was recently shown to constitute a very fitting correlate of spinal amplification phenomena underlying inflammatory pain [5] , [7] . This form of synaptic change has been reported to involve activation of NMDA receptors , NO release , and synthesis of cGMP [5] , [7] . However , which of the diverse targets of cGMP come into play at this synapse and how they mechanistically bring about long-lasting changes in the transfer of nociceptive information between the nociceptors and spinal neurons projecting to the brain is not understood so far . Furthermore , very little is known about exactly how neural circuits involved in pain processing are modulated by cGMP and which cellular and molecular processes underlie these changes . Studies on several different biological systems have shown that cGMP regulates multiple cellular targets , including diverse cGMP-gated ion channels , such as cyclic nucleotide-gated ( CNG ) and hyperpolarization-activated cyclic nucleotide-gated ( HCN ) channels , the cGMP-dependent protein kinases , PKG-I/cGK-I and PKG-II/cGK-II , as well as diverse phosphodiesterases ( PDEs ) [8] , [9] . Nearly all of these molecular targets of cGMP are expressed in nociceptive pathways and may potentially contribute to the key role of cGMP in synaptic potentiation in the spinal cord . Amongst these targets , PKG-I has emerged as a key mediator of cGMP functions in smooth muscle and platelet function [8] . The α-isoform of PKG-I has been reported to be expressed very highly in the primary sensory neurons in the dorsal root ganglia ( DRG ) over developmental [10] and adult stages [11] , and several regions in the brain and the spinal cord also express PKG-I [12] , [13] . Pharmacological and genetic studies in global , constitutive mutant mice have linked PKG-I to the development of the nociceptive circuitry as well as to spinal mechanisms of hyperalgesia [14] . Based upon this background , this study was designed with two goals in mind . First , it addressed the potential involvement of presynaptic mechanisms in the expression of synaptic potentiation on spinal projection neurons , which has not been explored or described previously . Second , it aimed to explore a potential role for PKG-I localized presynaptically in the spinal terminals of nociceptors in spinal potentiation and to clarify cellular and molecular mechanisms underlying these processes . We reasoned that the use of a conditional , region-specific gene deletion strategy to specifically manipulate presynaptic mechanisms might constitute an unambiguous approach towards addressing the above questions . Our results show that spinal synaptic potentiation triggered by nociceptor activation is associated with a long-lasting change in the probability of neurotransmitter release from spinal terminals of nociceptors . Using viable , developmentally normal transgenic mice lacking the PKG-I specifically in nociceptors with preserved expression in spinal neurons , brain , and all other organs , we demonstrate here that PKG-I localised in nociceptor terminals constitutes a key mediator of synaptic LTP and that its activation is functionally associated with pain hypersensitivity in vivo .
Mice lacking PKG-I specifically in a primary nociceptor-specific manner ( SNS-PKG-I−/− ) were generated via Cre/loxP-mediated recombination by mating mice carrying the floxed prkg1 allele ( PKG-Ifl/fl ) [15] with a mouse line expressing Cre recombinase under control of the Nav1 . 8 promoter ( SNS-Cre ) [16] . We have previously demonstrated that SNS-Cre mice enable gene recombination commencing at birth selectively in nociceptive ( Nav1 . 8-expressing ) sensory neurons , without affecting gene expression in the spinal cord , brain , or any other organs in the body [16] , [17] . An anti-PKG-I antibody [18] yielded specific staining in wild-type dorsal root ganglia ( DRG ) , but not in those from global PKG-I−/− mice [19] , thereby revealing Cre/loxP-mediated deletion of PKG-I in DRG of SNS-PKG-I−/− mice ( Figure 1A ) . Quantitative size-frequency analysis revealed that a majority of DRG neurons expressing PKG-I in wild-type mice are small-diameter neurons , which show a near complete loss of PKG-I expression in SNS-PKG-I−/− mice ( Figure 1B; p<0 . 001 ) . In contrast , a few large-diameter neurons showed low levels of anti-PKG-I immunoreactivity in DRGs of PKG-Ifl/fl , which was entirely retained in SNS-PKG-I−/− mice ( Figure 1A and B ) . Confocal analysis of dual immunofluorescence experiments revealed PKG-I immunoreactivity in nearly all Isolectin-B4 ( IB4 ) -labelled non-peptidergic nociceptors and substance P-expressing peptidergic nociceptors in PKG-Ifl/fl mice , both of which are selectively lost in SNS-PKG-I−/− mice ( typical examples in Figure 1C and quantitative summary in Figure 1D ) . In contrast , large-diameter neurofilament-200-immunoreactive neurons entirely retained PKG-I expression in the SNS-PKG-I−/− mice ( Figure 1C , D ) . Taken together , these results show that PKG-I is normally expressed in nearly all nociceptors and is selectively lost from these neurons , but not from tactile-responsive and proprioceptive DRG neurons , in SNS-PKG-I−/− mice . We found that PKG-I expression is entirely unaltered in the brains of SNS-PKG-I−/− mice ( an example of expression in cerebellar purkinje neurons is shown in Figure 1E , right panel ) , whereas global PKG-I−/− mice demonstrated a complete loss of anti-PKG-I immunoreactivity ( Figure 1E , right panel ) . In the spinal cord of SNS-PKG-I−/− mice , anti-PKG-I immunoreactivity was decreased selectively in the superficial dorsal laminae , which represent termination zones of the nociceptive afferents , as would be expected from SNS-Cre-mediated gene deletion in nociceptors ( Figure 1E , left panel ) . In contrast , neurons in the spinal cord entirely maintained immunoreactivity for PKG-I and appeared particularly conspicuous ( arrowheads in Figure 1E , left panel ) due to the loss of PKG-I labelling in afferent terminals in SNS-PKG-I−/− mice . Furthermore , anti-Cre immunohistochemistry as well as Western blot analysis with anti-PKG-I antibody confirmed that SNS-PKG-I−/− mice show a DRG-specific loss of PKG-I while retaining expression in the spinal cord and brain ( Figure S1 ) . In contrast to global PKG-I−/− mice , which typically demonstrate lethality in the first few weeks of life , SNS-PKG-I−/− mice were normal , fertile , and showed a normal life expectancy . In contrast to defects reported in global PKG-I−/− mice [10] , SNS-PKG-I−/− mice showed normal early targeting of TrkA-expressing primary afferents arising from the DRG ( arrowheads in Figure 1F , upper panels ) in the developing spinal dorsal horn at embryonic day 13 ( E13 ) . Unlike global PKG-I−/− mice [10] , SNS-PKG-I−/− mice did not show defects in T-branching of DiI-labelled primary afferents in the spinal cord over embryonic developmental stages ( arrows in Figure 1F , lower panels ) . Similarly , central and peripheral patterning of peptidergic or non-peptidergic nociceptors was normal in adult SNS-PKG-I−/−mice , as revealed by immunostaining for substance P and binding to IB4 , respectively , in the spinal dorsal horns and skin ( Figure S2A ) . Because peptidergic mechanisms have been suggested to play an important role in spinal LTP [5] , we ascertained that SNS-PKG-I−/− mice are not different from control mice with respect to the abundance of substance P in the spinal circuitry . Control and knockout mice exhibited the same prevalence of substance P-immunoreactive cells within DRG ( 33%±5% versus 29%±3% , respectively ) , which were not significantly different from each other ( p>0 . 05 , Student's t test ) . Moreover , the level of substance P immunoreactivity was similar in the superficial spinal dorsal horn across genotypes ( mean intensities in PKG-Ifl/fl mice and SNS-PKG-I−/− mice were 50±3 and 51±3 arbitrary units , respectively ) . Importantly , confocal microscopy revealed normal density of synapses between substance P-containing nociceptive afferents and PSD-95-positive puncta ( representing postsynaptic aspects of glutamatergic synapses ) in the spinal dorsal horns of SNS-PKG-I−/− mice as compared to PKG-Ifl/fl mice ( examples and quantification in Figure S2B ) . Finally , we addressed the internalization of NK1 receptors on spinal lamina I neurons following peripheral nociceptive stimulation in vivo , which has been demonstrated to be a clear indicator of nociceptive activity-induced synaptic release of substance P [20] . As shown in Figure S2C , application of a 52°C heat stimulus for 20 s to the plantar paw surface led to internalization of NK1 receptors in lamina I neurons of L3/L4 segments to a similar extent in SNS-PKG-I−/− and PKG-Ifl/fl mice ( quantification in Figure S2D ) . Unlike global PKG-I−/− mice [14] , SNS-PKG-I−/− mice showed a normal lamination of the spinal cord over early postnatal stages ( Figure S2E ) . Thus , the multiple developmental defects in the patterning of sensory afferents and spinal lamination that have been reported in global PKG-I−/− mice were not observed in SNS-PKG-I−/− mice . To address activity-dependent plasticity at spinal synapses , we recorded C-fiber-evoked synaptic LTP on spinal lamina I neurons projecting to the periaqueductal grey ( PAG ) , which were retrogradely labelled upon stereotactic injection of DiI in the PAG ( the experimental scheme is shown in Figure 2A and an example of a labelled cell is shown in Figure S3A ) [5] . In spinal-PAG projection neurons of wild-type mice , a conditioning low frequency stimulation of 2 Hz for 2 min produced synaptic LTP of monosynaptic C-fiber evoked EPSCs by more than 200% at 30 min ( Figure 2B ) . LTP at these synapses was preserved in the presence of strychnine and gabazine , which block glycinergic and GABAergic inhibitory neurotransmission , respectively ( Figure 2B ) . Similar results were obtained upon using another standard blocker of GABAergic neurotransmission , bicuculline , in combination with strychnine ( Figure S3B ) . Hence , LTP does not manifest due to primary afferent depolarization mediated by presynaptic GABA receptors or disinhibition of the postsynaptic neuron . To test whether LTP requires a postsynaptic function of PKG-I , we dialyzed standard PKG-I inhibitors , such as the non-permeant peptide inhibitor RKRARKE [21] , [22] or KT5823 [23] , into spinal neurons via the patch pipette . These manipulations did not affect the magnitude or duration of C-fiber-evoked LTP at spino-PAG synapses ( Figure 2C and Figure S3C ) , suggesting that PKG-I localized postsynaptically in spino-PAG projection neurons does not play a role in LTP at this synapse . To assess the role of PKG-I localized presynaptically in spinal nociceptor terminals , we then analysed PKG-Ifl/fl mice and SNS-PKG-I−/− mice . In spinal-PAG projection neurons of PKG-Ifl/fl mice , a conditioning low frequency stimulation of 2 Hz for 2 min produced LTP with a magnitude of more than 200% at 30 min and more than 300% by 60 min ( typical examples of time course and EPSC traces are given in Figure 2D , E ) . Prior to the conditioning stimulus , baseline values of C-fiber-evoked EPSCs stayed constant over the period of recording in both genotypes ( Figure 2D , E ) . In striking contrast to PKG-Ifl/fl mice , the conditioning stimulus did not evoke LTP in spinal-PAG projection neurons in SNS-PKG-I−/− mice ( Figure 2D , E; see Figure 2F and 2G for quantitative summary at 30 min post-conditioning stimulus; p<0 . 001; at least 13 neurons from each genotype were tested ) . For a clear interpretation of these data , it is imperative to address how basal nociceptive transmission at these synapses is affected in SNS-PKG-I−/− mice . Analysis of EPSC magnitude evoked by the first and the last pulse of the conditioning train revealed short-term depression of evoked EPSCs during the conditioning train , which was equivalent in PKG-Ifl/fl mice and SNS-PKG-I−/− mice ( Figure 3A; p = 0 . 95 ) , showing that the conditioning stimulus was equally effective in mice from both groups . Furthermore , basal C-fiber-evoked EPSCs were comparable between PKG-Ifl/fl and SNS-PKG-I−/− mice . We also established detailed input-output curves representing the relationship between the intensity of dorsal root stimulation and evoked EPSCs in the absence of a conditioning stimulus and found no differences between PKG-Ifl/fl and SNS-PKG-I−/− mice ( Figure 3B; p = 0 . 74 ) . Furthermore , the intensity of dorsal root stimulation required to elicit an action potential in post-synaptic spinal-PAG projection neurons was identical in PKG-Ifl/fl and SNS-PKG-I−/− mice ( an example is shown in Figure 3C ) . The intact nature of responsiveness in spinal-PAG projection neurons was demonstrated by similarities in their activation profiles upon current injection in SNS-PKG-I−/− mice and their PKG-Ifl/fl littermates ( an example is shown in Figure 3D ) . In all of the above experiments , quantitative analyses revealed that resting membrane potential , action potential width , delay after stimulation artefact , action potential threshold , delay for generation of first action potential ( latency to first AP ) , as well as amplitude of after hyperpolarisation ( AHP ) were similar in SNS-PKG-I−/− mice and PKG-Ifl/fl mice ( Figure 3E; p>0 . 05 in all cases , Student's t test ) . Finally , as an additional indicator of the number of fibers activated during electrical stimulation , we recorded fiber volleys in input/output measurements . Recording C-fiber volleys in the L4 and L5 dorsal roots derived from PKG-Ifl/fl and SNS-PKG-I−/− mice revealed typical responses , which increased in amplitude with increasing stimulus intensity ( representative traces are shown in Figure 3F ) . The amplitudes of C-fiber volley responses were not significantly different between PKG-Ifl/fl and SNS-PKG-I−/− mice ( stimulus-response curves are shown in Figure 3G; p>0 . 05; n = 16 per genotype ) , demonstrating directly that the number of fibers activated upon electrical stimulation was comparable between genotypes and could therefore not explain the failure to evoke synaptic potentiation in SNS-PKG-I−/− mice . These comprehensive analyses show that a presynaptic loss of PKG-I was specifically linked to a failure of activity-dependent potentiation of transmission at synapses between nociceptors and spinal-PAG projection neurons , but not to modulation of basal synaptic transmission . Following our observation that a specific presynaptic alteration in PKG-I expression perturbed spinal LTP , we then addressed potential contributions of presynaptic mechanisms at synapses between nociceptors and spinal projection neurons [7] , [24] . By recording miniature EPSCs in spinal-PAG projection neurons of control mice , we observed that quantal content varies largely , which is expected since these spinal projection neurons receive multisynaptic inputs from primary afferents as well as spinal interneurons . Furthermore , inputs arising from spinal interneurons are not expected to change in SNS-PKG-I−/− mice since the molecular perturbation is specific to nociceptor terminals . Thus , the net contribution of C-fibers to the population of mEPSCs is difficult to assess because it is unclear which fraction of mEPSCs can be attributed to C-fibers , which then makes detection of potentially small presynaptic changes highly unlikely when performing mini-analysis . To study synaptic events which could be clearly assigned to activation of presynaptic primary afferent fibers alone , we employed a protocol of minimal stimulation , setting the dorsal root stimulation parameters such that a synaptic failure rate of approximately 60% was achieved in recording solution containing 1 mM Ca2+ and 5 mM Mg2+ . The failure rate remained constant over a period of at least 30 min upon repetitive test stimulation in the absence of a conditioning stimulus ( 55 . 9%±3 . 9% pre- and 57 . 1%±8 . 1% at 30 min ) . However , upon application of the conditioning stimulus , minimal stimulation using the same parameters evoked a decrease in the frequency of synaptic failures within a few minutes in slices derived from PKG-Ifl/fl mice , indicating a change in the probability of neurotransmitter release; decrease in synaptic failures was accompanied by a corresponding rise in the magnitude of C-fiber-evoked EPSCs ( see Figure 4A for typical example and Figure 4C for quantitative summary of C-fiber-evoked EPSCs recorded every 15 s; n = 5; p = 0 . 04 ) . In contrast , the rate of synaptic failures did not change significantly following conditioning stimulus in slices derived from SNS-PKG-I−/− mice ( see Figure 4B and 4C; n = 7; p = 0 . 68 ) . EPSC values at the end of the recording were not significantly elevated as compared to basal values in SNS-PKG-I−/− mice ( 25±1 . 7 pA before and 21 . 1±1 . 2 pA at 30 min after the conditioning stimulus; p = 0 . 14 ) . In analogy to studies on hippocampal circuits , although the above evidence for a decrease in failure rate is indicative of increased presynaptic release probability , it has also been linked to unsilencing of silent synapses [25] . Therefore , to further consolidate presynaptic mechanisms , we focussed on the analysis of paired-pulse facilitation ( PPF ) , which represents a short-lasting increase in the second evoked EPSP when it follows shortly after the first and is well accepted as an indication of presynaptic mechanisms of long-term potentiation in the hippocampus [26] . In hippocampal CA1 neurons , PPF can increase as well as decrease in conjunction with LTP in a manner inversely proportional to the PPF prior to the conditioning stimulus [26] . Indeed , we obtained similar results in recordings at spinal synapses between C-fibers and spinal-PAG projection neurons . In spinal slices derived from mice of both genotypes , we found evidence for PPF as well as paired-pulse depression ( PPD ) prior to the LTP-inducing conditioning stimulus ( typical traces are shown in Figure 4D ) . Whereas a majority of neurons derived from PKG-Ifl/fl mice demonstrated a clear change in PPF or PPD following conditioning stimulus , neurons derived from SNS-PKG-I−/− mice did not ( see examples in Figure 4D ) . We then plotted the paired-pulse ratio ( PPR ) of the entire cohort of recorded neurons at 30 min after conditioning stimulation as a function of the basal PPR recorded prior to the conditioning stimulus ( Figure 4E , F ) . This analysis revealed that neurons in PKG-Ifl/fl mice with larger basal values of PPR prior to the conditioning stimulus ( indicated by filled round symbols in Figure 4E ) consistently showed a decrease in PPR after the conditioning stimulus ( Figure 4D ) , which is indicative of an increase in the probability of release ( PR , Figure 4E ) . This drop in PPF following conditioning stimulation did not come about or was reduced in neurons from SNS-PKG-I−/− mice ( filled round symbols in Figure 4F ) . A smaller cohort of synapses in PKG-Ifl/fl mice showed an increase in PPF after conditioning stimulation , but this was restricted to neurons with a low magnitude of PPF prior to the conditioning stimulus ( i . e . , a PPR of about 1 . 1–1 . 2 , filled square symbols in Figure 4E ) and a low expression of LTP ( Figure S4A ) . Again , this change was not observed in the corresponding cohort of neurons in SNS-PKG-I−/− mice ( filled square symbols in Figure 4F , Figure S4B ) . Conversely , in PKG-Ifl/flmice , higher magnitudes of LTP ( i . e . , between 150% and 350% , indicated by black frame in Figure S4A ) were consistently associated with a decrease in PPR , which is indicative of an increase in release probability . Neither LTP nor consistent changes in PPR were observed in SNS-PKG-I−/− mice ( Figure S4B ) . In conclusion , the failure rate analysis and PPR analysis strongly support the inference that the expression of LTP at spino-PAG synapses comes about via presynaptic mechanisms involving an increase in release probability via PKG-I . In an effort to understand the underlying molecular mechanisms , we then addressed potential substrates for the kinase activity of PKG-I in nociceptors . In particular , we reviewed known substrates of PKG-I in other biological systems and focussed on those for which we hypothesized a role in synaptic transmission . We first set up an assay system for testing involvement of PKG-I substrates in the DRG selectively upon persistent nociceptive stimulation in vivo , using Vasodilator-stimulated phosphoprotein ( VASP ) , a classical target of PKG-I , as an indicator of PKG-I activity [8] , [27] . Lysates of L4-L5 DRGs from naïve PKG-Ifl/fl and SNS-PKG-I−/− mice showed comparable levels of VASP expression ( Figure S5A , basal ) . Within minutes after persistent nociceptive stimulation via injection of formalin in the hindpaw , L4-L5 DRGs from PKG-Ifl/fl mice showed a striking phosphorylation of VASP at Serine 239 ( typical examples in Figure S5A and summary in Figure S5B , C; see Text S1 for details; p = 0 . 03 as compared to basal ) . This was markedly reduced in formalin-injected SNS-PKG-I−/− mice ( Figure S5A , B; p = 0 . 18 as compared to basal SNS-PKG-I−/− mice and 0 . 03 as compared to formalin-injected PKG-Ifl/fl mice ) . These results show that persistent activation of nociceptors leads to rapid signalling via PKG-I in DRG neurons in vivo , which is lost in SNS-PKG-I−/− mice , as expected . Using this assay system , we then addressed another key target of PKG-I , which has been mainly studied so far mechanistically in smooth muscle cells . Dephosphorylation of myosin light chains ( MLC ) [28] via PKG-I-dependent phosphorylation and activation of myosin light chain phosphatase in smooth muscle cells is a decisive mechanism underlying NO-mediated vasodilation [8] . Following formalin injection in the paw , we observed a strong phosphorylation of MLC in L4-L5 DRGs from PKG-Ifl/fl mice , which was found to be lacking in formalin-treated SNS-PKG-I−/− mice ( Figure 5A and 5B ) . These differences did not arise due to differences in expression levels of MLC between SNS-PKG-I−/− mice and PKG-Ifl/fl mice ( Figure 5B; Figure S5C ) . This finding was unexpected because it suggests a role for PKG-I in increasing MLC phosphorylation in DRG neurons , which is contrary to the classical role ascribed to PKG-I in MLC dephosphorylation . We reasoned that if our findings hold true , synthesis of cGMP ought to be a critical intermediate step in activity-dependent MLC phosphorylation in DRG neurons . Indeed , in mice pre-treated with an inhibitor of the soluble guanylyl cyclase , ODQ , and a pan-inhibitor of membrane-bound guanylyl cyclases , LY83583 , via intrathecal application , formalin-induced MLC phosphorylation in L4-L5 DRGs was strongly reduced ( see examples in Figure 5C; quantitative summary from three experiments is given below the Western blot ) . Immunohistochemistry revealed that nociceptor activation-induced increase in MLC phosphorylation occurred in the spinal termination zone of nociceptors ( lamina I and II ) as well as in spinal neurons ( Figure 5D ) . Interestingly , synaptic potentiation induced by a conditioning stimulus on spinal-PAG projection neurons was abolished in the presence of ML-7 , an inhibitor of MLCK ( Figure 5E and Figure 5F; p = 0 . 004 as compared to vehicle-treated control slices ) . Furthermore , consistent with our observations in SNS-PKG-I−/− mice , inhibition of MLC phosphorylation did not affect basal transmission at this synapse ( Figure 5G; p = 0 . 761 ) . Thus , the PKG-I target , pMLC , is functionally linked to potentiation of synaptic transmission in nociceptive laminae . Ikeda et al . [5] have reported that inhibition of IP3R activation blocks conditioning stimulus-induced synaptic potentiation at synapses between nociceptors and spinal-PAG projection neurons . This is particularly interesting because IP3R1 contains a PKG-I-recognition motif at serine 1755 and has been reported to be phosphorylated by PKG-I in vitro , putatively leading to gain of function [29] , [30] . We observed that IP3R1 is indeed a target of PKG-I in nociceptors and is functionally associated with modulation of calcium release from intracellular stores . In immunoprecipitation experiments from L4-L5 DRGs , formalin-injected PKG-Ifl/fl mice demonstrated highly enhanced serine 1755 phosphorylation of IP3R1 over the basal state; this effect was markedly reduced in DRGs obtained from formalin-injected SNS-PKG-I−/− mice ( Figure 6A , B ) , although expression levels of IP3R1 were comparable between SNS-PKG-I−/− mice and PKG-Ifl/fl mice ( Figure S5D ) . PKG-I-mediated phosphorylation of serine 1755 in IP3R1 has been suggested to positively modulate IP3R1 activity in heterologous test systems [30] . We observed that this function of PKG-I indeed plays an important role in modulating calcium release from intracellular stores in nociceptive neurons of the DRG . We performed Fura-2-based calcium imaging on dissociated DRG neurons derived from PKG-Ifl/fl and SNS-PKG-I−/− mice using Fluro488-conjugated Isolectin B4 ( IB4-Fluro488 ) for live identification of small-diameter nociceptive neurons ( neurons dually labelled with Fura-2 and IB4-Fluor488 are indicated by arrowheads in Figure 6C ) . The baseline values of the Fura2 ratios ( F340/F380 ) were not significantly different between control mice ( 1 . 049±0 . 010 ) and SNS-PKG-I−/− mice ( 1 . 040±0 . 008 ) ( p>0 . 05; Student's t test ) . Stimulation of calcium release via activation of Gq/11-phospholipase C-IP3R1 pathway by addition of ligands , such as bradykinin ( BK ) and the P2Y-receptor ligand , UTP , led to typical increases in the ratio of Fura-2 fluorescence at 340/380 nm in neurons from PKG-Ifl/fl mice , which were markedly reduced in neurons from SNS-PKG-I−/− mice ( see Figure 6D for typical examples and Figure 6E for quantitative summary; p<0 . 001 with respect to BK and UTP ) . In contrast , Fura2-labelled neurons with large-diameter somata , which were IB4-negative , did not show differences in calcium responses between PKG-Ifl/fl mice and SNS-PKG-I−/− mice ( an example is indicated by arrow in Figure 6C and quantitative summary is given in Figure 6E ) . In contrast , rapid calcium influx caused by KCl-induced depolarisation or capsaicin-evoked influx of calcium via TRPV channels was comparable in DRG neurons derived from SNS-PKG-I−/− mice and PKG-Ifl/fl mice ( Figure 6D , E; p>0 . 05 ) , showing thereby that a loss of PKG-I in nociceptive neurons is specifically linked to defects in IP3R-mediated calcium release from intracellular stores . Taken together , these biochemical and functional experiments suggest that following persistent nociceptive stimulation , PKG-I mediates potentiation of IP3R1 activity and MLC phosphorylation in sensory neurons , which is functionally linked to synaptic LTP at synapses between C-nociceptors and spinal-PAG projection neurons . We then went on to address whether these findings bear relevance to pain-related behaviour in vivo and found a functional role for PKG-I and its substrates in behavioural paradigms for spinal sensitization . As a test system , we studied the Phase II of formalin-induced nocifensive behavioural responses , which are manifest at 10–60 min after intraplantar formalin injection , for two reasons: one , this represents a widely used paradigm for studying central changes in pain processing caused by a persistent activation of nociceptors [31] , and two , intraplantar formalin induces synaptic LTP on spinal projection neurons with a matching time-course [5] . Formalin-induced phase II responses were significantly reduced upon intrathecal pretreatment with 2-APB or ML-7 to the lumbar spinal cord ( Figure 7A; p<0 . 01 for 2-APB and ML-7 in comparison to vehicle control , respectively ) , implicating involvement of IP3R function and MLC phosphorylation , respectively . Similarly , SNS-PKG-I−/− mice showed markedly reduced phase II responses than PKG-Ifl/fl mice ( Figure 7B; p<0 . 001 as compared to PKG-Ifl/fl mice ) . Basal withdrawal thresholds and response latencies to acute application of paw pressure ( e . g . , as tested with a dynamic aesthesiometer ) ( Figure S6A , left panel ) or thermal stimuli ( e . g . , a radiant infrared heat ramp ) ( Figure S6A , right panel ) , respectively , to the paw surface were found to be similar across SNS-PKG-I−/− mice and their control littermates ( p>0 . 05 ) . Furthermore , motor performance on a Rotarod was unaffected in SNS-PKG-I−/− mice ( Figure S6B; p = 0 . 20 ) . We have previously shown in details that SNS-Cre mice show no alterations in the processing of acute pain or chronic inflammatory or neuropathic pain [6] , [32] . In the context of studying disease-induced pain hypersensitivity , we first focussed on a model of inflammatory pain which is associated with primary hyperalgesia in the inflamed area and ongoing nociceptive inputs from the periphery throughout the time of testing , namely unilateral hindpaw inflammation induced by injection of Complete Freund's Adjuvant ( CFA ) [32] , [33] . CFA injection produced similar levels of edema in SNS-PKG-I−/− and PKG-Ifl/fl mice ( Figure S6C ) and hypersensitivity to graded von Frey mechanical stimuli ( Figure 7C , D ) or to plantar heat ( Figure 7E ) applied to the ipsilateral paw was assessed at 6 , 12 , 24 , 48 , and 96 h thereafter . Following inflammation , PKG-Ifl/fl mice demonstrated the characteristic leftward and upward shift in the stimulus-response curve over basal curves reflecting mechanical hypersensitivity ( black squares in Figure 7C ) . In contrast , SNS-PKG-I−/− mice demonstrated a less marked deviation from baseline behaviour upon CFA-induced inflammation ( red squares in Figure 7C ) . Furthermore , the relative drop in response thresholds to von Frey hairs ( defined here as minimum force required to elicit 40% response frequency ) in the inflamed state over basal ( pre-CFA ) state occurred to a significantly lesser extent in SNS-PKG-I−/− mice as compared to PKG-Ifl/fl mice ( left panel in Figure 7D; p<0 . 05 at all time points tested ) . Finally , SNS-PKG-I−/− mice showed a significantly lower magnitude of thermal hyperalgesia than PKG-Ifl/fl mice at 6 h after CFA and did not show hyperalgesia at all from 12 h onwards after CFA injection , whereas PKG-Ifl/fl mice continued to show thermal hyperalgesia all the way up to the latest time point tested , namely 96 h following CFA injection ( Figure 7E; p<0 . 01 between PKG-Ifl/fl and SNS-PKG-I−/− mice at all time points tested ) . We infer from the above that the development of primary hyperalgesia and mechanical allodynia following somatic inflammation is impaired by a loss of PKG-I in nociceptors . Although perturbation of spinal LTP may have contributed to the above phenotype in SNS-PKG-I−/− mice , it is conceivable that a peripheral role for PKG-I in nociceptors may at least partially account for changes in primary hyperalgesia . To address functional changes in nociceptor sensitivity in the inflamed tissue , we utilised the skin-nerve preparation [34] to study the electrophysiological properties of identified polymodal C-fibres and Aδ-mechanoceptors ( AM ) in the saphenous nerve . The excitability of mechanoreceptive C-fibers and AM-fibers showed a small , but significant , increase following paw inflammation in PKG-Ifl/fl mice ( see Figure 7F for typical examples ) , but not in SNS-PKG-I−/− mice ( Figure 7F ) . These data indicate defects in the development of peripheral sensitization in nociceptors of SNS-PKG-I−/− mice , which could contribute to a reduction in primary hyperalgesia; however , they are unlikely to account for the marked defects in mechanical allodynia observed following inflammation in SNS-PKG-I−/− mice . To explore central contributions , we utilised two models of aberrant pain which are triggered initially by peripheral inputs but do not require ongoing nociceptor activity in the periphery for maintenance . For example , capsaicin injection in the skin activates C-fibers and evokes hyperalgesia in the area of the flare ( primary hyperalegsia ) as well as outside of the flare ( secondary hyperalgesia ) . In PKG-Ifl/fl mice , we observed that injection of capsaicin in the skin of the lower thigh produced a marked allodynia at the hindpaw plantar surface , which was clearly excluded from the area capsaicin-induced flare ( see shift in von Frey response frequency in Figure 8A; black symbols ) . SNS-PGK-I−/− mice showed markedly reduced secondary hypersensitivity with capsaicin as compared to PKG-Ifl/fl mice ( red symbols in Figure 8A ) . Moreover , a capsaicin-induced drop in mechanical threshold ( allodynia ) was markedly reduced in SNS-PGK-I−/− mice as compared to PKG-Ifl/fl mice ( Figure 8B ) . It is well accepted that capsaicin-induced secondary mechanical hypersensitivity reflects C-fiber-evoked central amplification processes and can last for several hours , long after nociceptor responses to capsaicin have ceased owing to desensitisation of TRP channels [35] . Nevertheless , to rule out a potential contribution of ongoing peripheral inputs to the above-described phenotypic differences , we performed experiments in which nerve conduction was blocked with lidocaine in the peripheral dermatome in which capsaicin was injected in wild-type mice . As expected , lidocaine-induced nerve blockade prior to capsaicin injection blocked the induction of capsaicin-induced mechanical hypersensitivity ( Figure 8C ) ; in contrast , when lidocaine was injected 15 min after capsaicin , mechanical hypersensitivity developed normally ( Figure 8C ) , indicating that beyond the initial trigger , capsaicin-induced mechanical hypersensitivity is independent of ongoing input from peripheral nociceptors . In further experiments , we addressed the peripheral and central contributions of PKG-I . Pharmacological inhibition of PKG-I with KT5823 injected prior to injection of capsaicin in the same dermatome in wild-type mice did not block the development of capsaicin-induced mechanical hypersensitivity ( Figure 8D ) ; in contrast , when KT5823 was injected intrathecally prior to peripheral capsaicin injection , the induction of mechanical hypersensitivity was markedly inhibited ( Figure 8E ) , indicating a role for central PKG-I , but not peripherally expressed PKG-I . To further delineate the origin of the central ( spinal ) locus of PKG-I function , we undertook similar experiments in PKG-Ifl/fl mice and SNS-PGK-I−/− mice . Interestingly , intrathecally administered PKG-I inhibitor blocked the development of capsaicin-induced mechanical hypersensitivity in PKG-Ifl/fl mice and did not lower mechanical sensitivity any further in SNS-PGK-I−/− mice ( Figure 8F ) , demonstrating thereby the presynaptic locus of its action . In the muscle pain model by Sluka and colleagues [36] , two consecutive injections of dilute acidic saline in the flank muscle lead to secondary mechanical hypersensitivity in the ipsilateral and contralateral paws , which lasts for several weeks . The initial peripheral insult ( i . e . , flank muscle ) is spatially distinct from the area of application of nociceptive stimuli ( paw surface ) , ruling out a contribution of peripheral paw sensitization to the behavioural phenotype . Secondly , it has been shown in details previously that the secondary hyperalgesia in the paw lasts for several months after muscle injection , is not associated with any persistent inflammation or injury to the muscle tissue , is independent of peripheral inputs , and is thus central in origin [36] . Upon testing at 24 h after the induction of muscle pain , PKG-Ifl/fl mice demonstrated a pronounced leftward and upward shift in the stimulus-response curve to von Frey hairs applied to the plantar paw surface ( black squares in Figure 9A , middle panel ) , which was still evident 3 wk later in the ipsilateral ( Figure 9A , right panel ) ; this changes come about in the paw ipsilateral to the injected flank muscle ( upper panels in Figure 9A ) as well as in the contralateral paw ( lower panels in Figure 9A ) . In contrast , SNS-PKG-I−/− mice did not show significant deviations ( red squares in Figure 9A ) . Analysis of paw-withdrawal thresholds to graded pressure also consistently revealed that muscle injection-induced drop in paw mechanical thresholds at the paws was significantly lesser in SNS-PKG-I−/− mice than in control littermates at all time points tested ( Figure 9B ) . In conclusion , these analyses support an essential role for presynaptic PKG-I in nociceptor terminals in central mechanisms of secondary mechanical hypersensitivity . Finally , we asked whether PKG-I expressed in nociceptors constitutes an important target of the NMDA-NOS-cGMP pathway . Consistent with previous studies [37] , intrathecally administered NMDA produced a rapid facilitation of the paw withdrawal reflex ( Figure 10A ) . Importantly , in striking contrast to PKG-Ifl/fl mice ( black symbols ) , SNS-PKG-I−/− mice completely failed to develop hyperalgesia following intrathecal NMDA delivery ( red symbols , upper panel in Figure 10A ) . Similar results were obtained upon delivery of an NO donor , NOC-12 , to the spinal cord via intrathecal catheters ( upper panel in Figure 10B ) . Furthermore , intrathecal delivery of NMDA and NOC-12 produced a facilitation of the tail flick reflex in PKG-Ifl/fl mice , but not in SNS-PKG-I−/− mice ( lower panels in Figure 10A , B ) , showing thereby that PKG-I is critically required for the pro-nociceptive functions of the NMDA and NO . Because soluble guanylyl cyclases ( sGC ) represent a key molecular link between NO and activation of PKG-I , the above results imply that NO activates sGC in spinal presynaptic terminals of nociceptors . While some studies report a lack of sGC expression in DRG neurons [38] , others reported expression in a population of small diameter DRG neurons and in primary afferents [39] , [40] . Here , we carried out mRNA in situ hybridisation using riboprobes recognising the beta subunit of sGC on mouse DRG sections and observed distinct , specific signals over the soma of several large and small-diameter DRG neurons ( arrowheads and arrows in Figure 10C , respectively ) . Furthermore , the satellite cells surrounding DRG neurons showed dense signals ( red arrows in Figure 10C ) . Sense control probes did not yield any appreciable signals ( Figure 10C ) . These results indicate that sGC mRNA is expressed in sensory neurons of the DRG . In addition to sGC enzymes , which are directly activated upon NO , the membrane-bound guanylyl cyclases ( mGC; Npr family ) also contribute to cGMP production in some organs ( e . g . , in the cardiovascular system ) [41] . Stimulated by recent reports on expression of mGCs in DRG neurons [42] , we administered a cocktail of natriuretic peptides ( ANP , BNP , and CNP ) intrathecally and observed marked hyperalgesia within 15 min after delivery , which lasted for about 45–50 min in wild-type mice ( unpublished data ) and PKG-Ifl/fl mice ( Figure 10D ) . Interestingly , natriuretic peptide-induced hyperalgesia was also entirely abrogated in SNS-PKG-I−/− mice ( Figure 10D ) . These results suggest that the NMDA-NOS-soluble guanylyl cyclase-cGMP pathway as well as the natriuretic peptide-mGC-cGMP pathways converge upon PKG-I expressed in spinal terminals of nociceptors to modulate nociceptive processing in the spinal cord . Finally , we undertook experiments to test whether PKG-I expression alone or some downstream factor perturbed by an early loss of PKG-I is responsible for the deficits in pain hypersensitivity in SNS-PKG-I−/− mice . We constructed chimeric Adeno-associated virions of the serotypes AAV1 and AAV2 expressing an C-terminally GFP-tagged version of the murine PKG-I cDNA [43] . Injection in unilateral L3 and L4 DRGs in vivo led to a broad expression in the DRG . AAV1/2 chimeric virions expressing GFP alone served as controls . PKG-Ifl/fl mice and SNS-PKG-I−/− mice expressing GFP-tagged PKG-I or GFP alone showed normal basal sensitivity to graded von Frey stimuli ( Figure 11B ) . Upon peripheral injection of capsaicin , PKG-Ifl/fl mice expressing GFP-tagged PKG-I showed a small increase in mechanical hypersensitivity than PKG-Ifl/fl mice expressing GFP , which was only statistically significant at some intensities of mechanical stimuli ( Figure 11C ) . As expected , SNS-PKG-I−/− mice overexpressing GFP in DRG showed markedly reduced mechanical hypersensitivity with capsaicin than PKG-Ifl/fl mice overexpressing GFP . Importantly , overexpression of GFP-tagged PKG-I fully restored mechanical hypersensitivity in SNS-PKG-I−/− mice ( Figure 11D ) . This indicates that expression of PKG-I is both necessary and sufficient for inducing centrally maintained hypersensitivity upon persistent peripheral activation of C-fibers .
In contrast to the intensively studied forms of LTP in the hippocampus , very few studies have addressed cellular and molecular mechanisms of LTP at spinal synapses regulating the flow of nociceptive information from the periphery towards the brain [7] , [24] . Here we observed that nociceptive activity-driven LTP at synapses between nociceptive terminals and spinal neurons projecting nociceptive inputs to the PAG requires presynaptic mechanisms for its full expression . Furthermore , our results indicate that this function is mediated by cGMP acting via PKG-I . We base our inferences on three main observations: ( 1 ) A specific loss of PKG-I in presynaptic , but not post-synaptic , compartments of this synapse abolished C-fiber-evoked LTP without altering basal neurotransmission at this synapse; ( 2 ) LTP was temporally accompanied by a decrease in the rate of synaptic failures in a presynaptic-PKG-I-dependent manner; and ( 3 ) LTP was associated with a change in the PPR , which did not take place when PKG-I was deleted presynaptically in nociceptor terminals . Importantly , higher magnitudes of LTP were consistently associated with a decrease in PPF , and thereby with an increase in release probability . Previous studies have shown that the NMDA receptor-NO-cGMP pathway is important in the induction of spinal LTP [4] , [5] , and it has been assumed that this pathway comes into play in the post-synaptic compartment . However , all of the above signal transducers are also expressed presynaptically in afferent terminals in the spinal dorsal horn [44] . Thus , pre- and post-synaptic contributions to spinal LTP have not been worked out so far . There is evidence for a requirement for post-synaptic Ca2+ change for the induction of the LTP ( i . e . , in experiments with BAPTA in recording pipette; [5] ) . Taken together with our results , this suggests that a calcium-dependent postsynaptic mechanism may be required for the induction of LTP ( e . g . , via NMDA receptor-dependent generation of NO ) ; in contrast , a presynaptic change involving cGMP- and PKG-I-dependent increase in neurotransmitter release may mediate the expression of LTP at synapses between nociceptors and spinal-PAG projection neurons . Mechanistically , this may come about via involvement of multiple phosphorylation targets of PKG-I . While some targets have been identified in heterologous systems , very little is known about the nature and functional role of PKG-I targets in vivo . Here , we identified and validated two primary targets in DRG neurons , namely the IP3R1 and MLC , and observed that PKG-I modulates intracellular calcium release as well as MLC phosphorylation differently in DRG neurons as compared to other biological systems , such as the smooth muscle . For example , in some biological systems , PKG-I has been reported to negatively modulate calcium signals via its interaction with IRAG [18] . However , IRAG interacts selectively with the beta-isoform of PKG-I , which is barely expressed in the DRG , but not with PKG-I-alpha , the predominant form found in DRG neurons [10] . Furthermore , our observations that PKG-I potentiates calcium release induced by typical mediators of nociceptive sensitization , such as bradykinin , in identified nociceptive neurons and that repetitive activation of nociceptors in vivo leads to PKG-I-mediated phosphorylation of IP3R1 at serine 1755 , a site associated with positive functional modulation , implicate PKG-I as a positive modulator of calcium signalling in nociceptive neurons . In light of electrophysiological analyses reported here , this raises the possibility that calcium released from IP3R1-gated stores may participate in modulating presynaptic function . Although a few studies at hippocampal synapses have proposed an involvement of calcium stores in modulation of presynaptic release [45] , [46] , underlying cellular mechanisms are not known . Results of this study suggest that activation of presynaptic PKG-I may constitute the molecular link between synaptic activity and the elevation of resting levels of calcium in presynaptic terminals , thereby potentiating synaptic transmission via an increase in release probability . Furthermore , we observed that MLC was phosphorylated in a nociceptive activity-dependent manner and that PKG-I is required for MLC phosphorylation in the DRG . Although our electrophysiological data implicate phosphorylated MLC in LTP at spinal synapses , not much can be inferred about downstream mechanisms at this stage . At central synapses , MLC phosphorylation was initially implicated in vesicle transport and in regulation of vesicular pools [47]; however , these inferences could not be corroborated in detailed subsequent analyses [48] . Taken together , more detailed analyses overcoming current technical hindrances in studying mobilization of vesicular pools in the complex circuitry of the DRG and spinal dorsal horn will be required to understand mechanisms underlying PKG-I-mediated modulation of release probability at this synapse . The possibility of functionally linking synaptic changes described here to changes in nociceptive behaviour simultaneously represents a good opportunity and a major challenge . As the first parameter to test this relationship , we focused on the phase II responses in the intraplantar formalin test , which has been attributed to spinal nociceptive sensitization triggered by an initial barrage of C-fiber inputs [31] . Indeed , we observed that presynaptic loss of PKG-I as well as functional perturbation of MLCK and IP3R , its substrates involved in LTP , inhibited phase II behavioural responses . However , a contribution of central mechanisms could not be inferred from the formalin data due to several reasons: Although MLCK/IP3R inhibitors were administered spinally , the genetically induced loss of PKG-I in SNS mice occurred throughout the nociceptor , including peripheral terminals . Moreover , there is still some ongoing activation of peripheral nociceptors in the phase II of the formalin response [31] . Indeed , our electrophysiological analyses in the CFA inflammatory pain model suggested that peripheral PKG-I may contribute to primary hyperalgesia . Therefore , we focused on pain models in which nociceptive hypersensitivity is triggered by peripheral nociceptors , but maintained via central mechanisms that outlast and are independent of peripheral inputs . One of these is the chronic muscle pain model in which injections of dilute acidic saline in the gastronemius muscle evokes a long-lasting secondary mechanical hyperalgesia in the ipsilateral and contralateral paws , which lasts for several weeks , is unrelated to muscle damage and is not maintained by continued primary afferent input from the site of injury , as shown by experiments involving dorsal rhizotomy and lidocaine injections in the muscle [36] . Similarly , we addressed capsaicin-induced secondary mechanical hyperalgesia outside of the primary flare , which albeit triggered by C-fiber inputs , is maintained via mechanisms of central origin as indicated by previous studies [35] and our analyses . In both models , we found marked defects in central hypersensitivity in SNS-PKG-I−/− mice . Our results indicated that capsaicin-evoked mechanical hypersensitivity is neither dependent on peripheral PKG-I function nor does it require ongoing peripheral nociceptor sensitisation . Moreover , they revealed that PKG-I expressed in central terminals of nociceptors plays a decisive role in the induction of mechanical hypersensitivity after persistent C-fiber stimulation via capsaicin . Finally , reinstating PKG-I expression in the DRG in adult SNS-PKG mice fully restored capsaicin-evoked mechanical hypersensitivity , indicating that PKG-I directly , and not some factor affected by a genetic loss of PKG-I , is a functional determinant of C-fiber-evoked mechanical hypersensitivity . In summary , this study shows that PKG-I expressed in nociceptors terminals is the principal target of cGMP at nociceptive synapses . Furthermore , it suggests that PKG-I-mediated presynaptic facilitation and LTP in spinal projection neurons is functionally involved in activity-dependent centrally mediated nociceptive hypersensitivity .
Homozygous mice carrying the flox allele of the mouse prkg1 gene , which encodes the cGMP dependent kinase 1 ( PKG-Ifl/fl ) [15] , have been described previously in detail . PKG-Ifl/fl mice were crossed with SNS-Cre mice [16] to obtain PKG-Ifl/fl;SNS-Cre+ mice ( referred to as SNS-PKG-I−/− mice in this article ) and PKG-Ifl/fl mice ( control littermates ) . Mice were crossed into the C57BL6 background for more than 8 generations . Mice lacking PKG-I globally ( PKG-I−/− mice ) have been described before . Only littermates were used in all experiments to control for background effects . Mice ( 14–18 d old ) were anesthetized with a mixture of Dormitor , Dormicum , and Fentanyl , and stereotactic injections of DiI into the PAG were carried out ( see Text S1 for details ) . After 2 to 3 d , transverse 350–450 µm thick spinal cord slices with dorsal roots attached were obtained and whole cell patch clamp recordings of identified DiI-positive neurons were performed as described in Text S1 . Test pulses of 0 . 1 ms with intensity of 3 mA were given at 30 s intervals to the dorsal root via a suction electrode . For studying the site of expression of synaptic potentiation , we used minimal stimulation in conditions of low release probability ( in mM: NaCl 127; KCl , 1 . 8; KH2PO4 , 1 . 2; Ca2+ 1 . 0; Mg2+ , 5; NaHCO3 , 26; glucose , 15; oxygenated with 95% O2 , 5% CO2; pH 7 . 4 ) . Dorsal root was stimulated at intensity of threshold to evoke EPSCs on DiI-labelled spino-PAG projection neuron . Under these conditions , the failure rate was 60 . 9%±6 . 3% ( n = 10 ) . To induce synaptic potentiation , low frequency stimulation ( conditioning stimulus , 2 Hz for 2 min ) was applied to dorsal root as a conditioning stimulus with the same intensity as the test stimulus [5] . The recording mode during conditioning stimulation was the same as that before and after conditioning stimulation . Neurons are voltage clamped at −70 mV . Because a suction electrode was utilized to stimulate the whole root , a suprathreshold stimulus was required to fully recruit C-fibers in the root [5] . Synaptic strength was quantified by assessing the peak amplitudes of EPSCs . The mean amplitude of 4–5 EPSCs evoked by test stimuli prior to conditioning stimulation served as a control . Significant changes from control were assessed by measuring the peak amplitudes of five consecutive EPSCs every 5 min after conditioning stimulation . Additional details are given in Text S1 . In some experiments , blockers of inhibitory neurotransmission , such as Gabazine ( 10 µM ) and Strychnine ( 1 µM ) , were added to the bath . In a subset of animals , paired-pulse stimuli with an inter-stimulus interval of 110 ms ( 0 . 1 ms pulse duration , 3 mA intensity , every 30 s ) were used ( see Text S1 for details ) . Paired-pulse ratio ( facilitation or depression ) of C-fiber-evoked EPSC was calculated as the amplitude of the second C-eEPSC divided by that of the first C-eEPSC in a pair . In a subset of experiments , PKG-I inhibitors such as KT5823 ( 10 µM ) or RKRARKE ( 250 µM ) were infused post-synaptically via the patch pipette . The following antibodies were used for Western blots and biochemical analyses: anti-IP3R1 , anti-pS1755 IP3R1 ( kind gift from Prof . Richard Wojcikiewicz ) , anti-VASP ( Alexis Biochemical ) , anti-MLC , anti-alpha tubulin ( Sigma ) , anti-pSer239 VASP , anti-pThr18/pSer19 MLC ( Cell Signaling technology ) , anti-PKG-I antibody [18] , secondary HRP labelled anti-rabbit ( Sigma Aldrich ) , or anti-mouse ( GK Healthcare UK Ltd . ) . The following antibodies were used for immunohistochemistry: phospho-ERK1/2 antibody ( Cell signalling ) , anti-Fos antibody ( Chemicon ) , anti-Isolectin B4 antibody ( vector laboratories ) , anti-Calcitonin gene related peptide antibody ( Immunostar ) , anti-Neurofilament 200 antibody ( Chemicon ) , anti-Substance P antibody ( Chemicon ) , anti-PKG-I antibody [18] , anti-PSD-95 antibody ( a gift from M . Watanabe ) and anti-TrkA antibody ( a kind gift from Prof . L . F . Reichardt ) , and anti-cre antibody ( Novagen ) . The soluble guanylyl cyclase inhibitor , ODQ ( 25 mg/kg body weight; sigma Aldrich ) , and the membrane guanylyl cyclase inhibitor , LY83583 ( 12 . 5 mg/kg body weight; sigma Aldrich ) , were dissolved in 50% DMSO and injected in a volume of 250 µl intraperitoneally . The following drugs were administered intrathecally in vivo: an inhibitor of MLCK ( ML-7; 15 nmol Alexis Biochemical , dissolved in 5% DMSO ) , an inhibitor of IP3R ( 2-APB; 2 nmol; Calbiochem ) , NMDA ( 100 fmol; Sigma Aldrich ) , the NO donor , NOC-12 ( 17 nmol; Sigma Aldrich ) , atrial natriuretic peptide , brain natriuretic peptide and c-type natriuretic peptide ( rANP1-28 , mBNP45 , and hCNP1-22; 330 pmol of each natriuretic peptide; American peptide company , Inc . , USA ) , and the PKG-I inhibitor KT5283 ( 200 pmoles ) . See Text S1 for details on intrathecal delivery . Mice were allowed to recover for 2 d after surgery , and only animals showing complete lack of motor abnormalities were used for further experiments . 5 µl of drugs were applied followed by flushing of the catheter with 10 µl of 0 . 9% saline . The following drugs were administered peripherally in the vicinity of the paw in experiments pertaining to capsaicin-induced mechanical hypersensitivity: KT5823 ( 200 pmoles ) and lidocaine ( 10 µl of 2% ) . A total of 17 PKG-Ifl/fl and 15 SNS-PKG-I−/− mice were used in the electrophysiological recordings of nerve activity . An ex vivo skin-nerve preparation was used to study the properties of mechanosensitive C- and A-δ afferent fibres which innervate the skin in the inflamed area 24 h following CFA inoculation ( 20 µl ) as described previously ( see Text S1 for details ) . All animal use procedures were in accordance with ethical guidelines imposed by the local governing body ( Regierungspräsidium Karlsruhe , Germany ) . All behavioural measurements were done in awake , unrestrained , age-matched mice of both sexes that were more than 3 mo old by individuals who were blinded to the genotype of the mice being analyzed ( see Text S1 for details ) . The open reading frame of mouse PKG-I fused C-terminally with GFP [43] or EGFP alone was cloned in an AAV expression construct , and chimeric AAV1/2 virions were generated using standard protocols . Virions were diluted 1∶2 with 20% mannitol and injected unilaterally into L3 and L4 DRGs ( 1 µl per DRG , or approx . 107 transfection units per DRG ) in deeply anesthetized mice as described in detail previously [49] . Mice were tested in behavioural tests 2 wk after injection . At the end of the experiment , mice were perfused as described above and expression of GFP was confirmed via fluorescence analysis . All data are presented as mean ± standard error of the mean ( S . E . M . ) . For comparisons of multiple groups , analysis of variance ( ANOVA ) for random measures was performed followed by post hoc Fisher's test to determine statistically significant differences . When comparing two groups that were studied in parallel , Student's t test was employed . Unless otherwise specified , the p values shown in the figures and text are derived from ANOVA and post hoc Fisher's test . p<0 . 05 was considered significant . | Pain is an important physiological function that protects our body from harm . Pain-sensing neurons , called nociceptors , transduce harmful stimuli into electrical signals and transmit this information to the brain via the spinal cord . When nociceptors are persistently activated , such as after injury , the connections they make with neurons in the spinal cord are altered in a process called synaptic long-term potentiation ( LTP ) . In this study , we examine the molecular and cellular mechanisms of LTP at synapses from nociceptors onto spinal neurons . We use multiple experimental approaches in mice , from genetic to behavioural , to show that this form of LTP involves presynaptic events that unfold in nociceptors when they are repetitively activated . In particular , an enzyme activated by the second messenger cGMP , referred to as Protein Kinase G-I , phosphorylates presynaptic proteins and increases the release of neurotransmitters from nociceptor endings in the spinal cord . When we genetically silence Protein Kinase G-I or block its activation in nociceptors , inflammatory pain is markedly reduced at the behavioural level . These results clarify basic mechanisms of pathological pain and pave the way for new therapeutic approaches . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology",
"neuroscience"
] | 2012 | Presynaptically Localized Cyclic GMP-Dependent Protein Kinase 1 Is a Key Determinant of Spinal Synaptic Potentiation and Pain Hypersensitivity |
The gut-to-brain axis exhibits significant control over motivated behavior . However , mechanisms supporting this communication are poorly understood . We reveal that a gut-based bariatric surgery chronically elevates systemic bile acids and attenuates cocaine-induced elevations in accumbal dopamine . Notably , this surgery reduces reward-related behavior and psychomotor sensitization to cocaine . Utilizing a knockout mouse model , we have determined that a main mediator of these post-operative effects is the Takeda G protein-coupled bile acid receptor ( TGR5 ) . Viral restoration of TGR5 in the nucleus accumbens of TGR5 knockout animals is sufficient to restore cocaine reward , centrally localizing this TGR5-mediated modulation . These findings define TGR5 and bile acid signaling as pharmacological targets for the treatment of cocaine abuse and reveal a novel mechanism of gut-to-brain communication .
Traditionally , bile acids have been viewed as detergents participating in the emulsification of ingested fats . It is becoming increasingly apparent , however , that bile acids also function as steroid hormones with targets in the intestine , liver , and brain [1–4] . Bile acids produced from cholesterol in the liver enter the proximal small intestine at the duodenum and are reabsorbed into hepatic portal circulation at the distal ileum , a segment of the small intestine densely populated by bile acid receptors and reuptake transporters . Bile diversion—a newly developed bariatric surgical procedure in mice—is capable of chronically elevating circulating bile acids beyond the enterohepatic bile pool through ligation of the common bile duct and anastomosis of the gallbladder to the ileum ( GB-IL ) ( Fig 1A ) [5] . In the control surgery , the gallbladder is anastomosed to the duodenum ( GB-D ) ( Fig 1A ) , restoring normal bile flow as well as circulating bile acid levels . Bile diversion was recently developed in mice to treat high-fat diet–induced obesity [5] . GB-IL mice exhibit reduced high fat food consumption as well as weight loss . This reduction in the intake of rewarding , calorically dense food could stem at least in part from altered valuation of palatable food . Reward is a process regulated , among other factors , by dopamine ( DA ) signaling and homeostasis . Dysregulated mesolimbic DA circuitry has been linked to augmented high-fat , high-calorie food consumption [6–8] and , importantly , to cocaine abuse [9–12] . We thus hypothesized that bile diversion to the ileum , which reduces hedonic feeding , might also reduce the rewarding properties of cocaine . Alteration in cocaine reward promoted by GB-IL would suggest a generalized mechanism by which bile acids regulate central encoding of rewards . Here , we show that GB-IL surgery is able to alter the behavioral and pharmacological responses to cocaine . This led us to uncover a novel role of central bile acid signaling mediated by Takeda G protein-coupled bile acid receptor ( TGR5 ) for cocaine-induced impairments in DA homeostasis and the development of associated behaviors .
Cocaine directly alters DA neurotransmission and produces its rewarding effects by increasing available extracellular DA in specific brain regions , including the nucleus accumbens ( NAc ) [6 , 13] . Behavioral pharmacological experiments indicate that increased DA transmission is clearly both necessary and sufficient to promote psychostimulant reinforcement , including the development of cocaine place preference ( CPP ) ( for review , see Pierce and Kumaresan [14] ) . To evaluate whether and how GB-IL surgery regulates the reinforcing properties of cocaine , we first studied its effect on cocaine’s ability to enhance electrically evoked DA release in NAc slices . We utilized animals in the final phase of CPP ( post-conditioning , day 10 , see Materials and methods section ) . Three stable baseline recordings were taken at five-minute stimulation intervals , and no differences were noted between GB-D ( Fig 1C , baseline ) and GB-IL ( Fig 1C , baseline ) in terms of peak amperometric current ( Fig 1D ) . Similar peak amperometric currents between the two conditions establishes that the presynaptic properties at DA synapses are unchanged by the GB-IL surgery . However , the increase in electrically evoked DA release promoted by 10 μM cocaine was significantly reduced in the GB-IL mice ( Fig 1C , cocaine; quantitation in Fig 1E as area under the curve [AUC] ) . To further determine whether multiple exposures to cocaine are required for the GB-IL surgery to reduce cocaine’s ability to augment DA release , amperometric experiments were performed in GB-D and GB-IL animals receiving the vehicle instead of cocaine as described in Fig 1C and 1D . In cocaine-naïve animals , the GB-IL surgery significantly impaired cocaine-induced increase in DA release without altering the peak of the amperometric current ( S1 Fig ) . These data indicate that cocaine has impaired ability to increase released DA in GB-IL mice independently of cocaine exposure . GB-IL mice ( post-conditioning , day 10 ) do not exhibit an overt neurochemical phenotype , as total accumbal tissue levels of DA and its related monoamines , norepinephrine ( NE ) and serotonin ( 5-HT ) , were not significantly altered with respect to GB-D ( Fig 1F ) . This result further suggests that the changes in DA homeostasis promoted by cocaine are not due to the supply or release properties of DA . Based on the alterations in their pharmacological response to cocaine , we next determined whether GB-IL mice display reduced behavioral responses to cocaine . Mice were tested for cocaine conditioned place preference ( CPP; 20 mg/kg , intraperitoneally [i . p . ] ) in a dual compartment apparatus with features allowing for the animals to distinguish between the two compartments ( Fig 2A ) . Prior to drug administration in the pre-test session , no differences in the level of cumulative baseline locomotion were observed over 30 minutes ( GB-IL was 85% ± 8% of GB-D; n = 11–14 per group; p > 0 . 2 by Student t test ) . Next , one compartment of the apparatus was paired with experimenter-administered cocaine , while the second compartment was paired with experimenter-administered saline . During conditioning sessions , locomotor behavior was collected . On first exposure to cocaine , cocaine-induced hyperlocomotion was indistinguishable between the two groups . Notably , while GB-D control mice exhibited significant locomotor sensitization to cocaine over multiple exposures , the GB-IL mice did not ( Fig 2B ) . Prior work strongly suggests that psychomotor sensitization is associated with the development of molecular adaptations within the mesocorticolimbic system in the development of an addiction [15] . The lack of locomotor sensitization in our bile diversion model may thus support impairments in the central encoding of cocaine reward . Importantly , while both groups formed a place preference for cocaine , the preference of GB-IL mice for the cocaine-paired side was significantly less than that observed for GB-D mice ( Fig 2C ) . Four to seven days following CPP , GB-D and GB-IL mice were tested for open-field ( OF ) locomotion . In an OF , neither spontaneous nor saline-induced locomotion in GB-IL mice significantly differed from GB-D mice; however , cocaine-induced locomotion ( 20 mg/kg , i . p . ) was significantly attenuated between 10 and 40 minutes post-injection ( Fig 2D and inset ) . No change in pre-test or OF spontaneous locomotion ensures that changes in compartment preference in the CPP task were not the result of reduced locomotion in the GB-IL group . These changes in the behavioral response to cocaine are also not the result of reduced cocaine bioavailability in the striatum . We measured striatal cocaine availability in GB-D and GB-IL mice by liquid chromatography mass spectrometry ( Mass Spectrometry Core , Vanderbilt University ) . In mice injected with cocaine ( 20 mg/kg , i . p . ) 30 minutes prior to being euthanized we did not detect any significant difference in striatal cocaine ( GB-IL was 104 ± 20% of GB-D; n = 9–5 per group; p > 0 . 8 by Student t test ) . Furthermore , the reduction in conditioning to cocaine cannot be attributed to impaired spatial learning or memory capabilities , as we did not observe any significant impairment in performance on a hidden water maze ( HWM ) task ( S2A–S2C Fig ) . No generalized impairments in motor abilities in a rotarod test ( S2D Fig ) or in a tail suspension test ( TST ) ( S2E Fig ) were observed . However , in the OF , we did observe a small but significant increase in center time in the GB-IL group ( S2F Fig ) , suggesting that the surgery may also affect systems regulating exploratory behavior or anxiety . Following cocaine exposure as per CPP , mice undergoing GB-IL surgery exhibit greatly increased blood levels of total and conjugated bile acids relative to GB-D , while levels of primary , secondary , and unconjugated bile acids remain unchanged ( Fig 3A ) . Such a dramatic increase of these nutrient-signaling hormones suggests a possible role for bile acids in producing the behavioral effects of the surgery . Moreover , this elevation points to a potent and previously unexplored role for bile acid signaling as a regulator of cocaine reward , which is the focus of this study . Notably , in control animals , administration of cocaine as described in Fig 2 did not significantly alter levels of total , conjugated , or unconjugated bile acids ( cocaine versus vehicle; data are expressed as percent of vehicle control; total bile acids 64 ± 11% , p > 0 . 079; conjugated bile acids 57 ± 14% , p > 0 . 079; unconjugated bile acids 99 ± 0 . 1% , p > 0 . 9 by Student t test , n = 7–8 per group ) . Since bile acid synthesis is regulated by the gut microbiota-to-liver axis [16] , we analyzed the relative abundance and distribution of the most highly abundant resolved bacterial families in GB-IL and GB-D fecal samples . We did not find any differences in the gut microbiome composition in our surgical models ( S3 Fig ) . Bile acids signal as hormones mainly through two bile acid receptors: the farnesoid x receptor ( NR1H4 , FXR ) and the G protein-coupled bile acid receptor 1 ( GPBAR1 , TGR5 ) , which is expressed in the brain [2 , 17] . Here , we show that chronic administration of the synthetic bile acid obeticholic acid ( OCA ) , an agonist of TGR5 as well as a potent agonist of FXR [18 , 19] , is sufficient to reduce cocaine CPP ( Fig 3B ) in wild-type mice . For two weeks prior to the initiation of cocaine CPP , mice were treated orally with OCA ( 10 mg/kg , Per os [p . o . ] ) or vehicle . The treatment continued for 4 weeks following drug initiation until euthanasia . Mice treated with OCA , compared with vehicle , exhibited decreased cocaine CPP ( Fig 3B ) . Several conjugated and unconjugated bile acids have been found to promote phosphorylation of extracellular signal-regulated kinase ( ERK ) 1/2 [20] . We determined that ERK 1/2 phosphorylation measured by immunochemistry as previously described [21] was significantly elevated in the NAc of mice chronically treated as in Fig 3B with OCA ( S4 Fig ) . These data suggest that oral administration with OCA signals at least in part in the NAc . Finally , we determined whether venous infusion of the bile acid tracer , 2 , 2 , 4 , 4-[2H]-taurocholic acid ( d4-TCA , 0 . 0038 μmol·kg-1·min-1 ) , administered at 1 . 5 μL/minute was capable of reaching the brain . After a 90-min tracer equilibration period , mice were anesthetized with an infusion of sodium pentobarbital , the brain was excised , immediately frozen in liquid nitrogen , and stored at −80 °C until analyzed . Mass spectrometry analysis was performed to calculate the levels of taurocholic acid-d4 ( d4-TCA ) in brain samples ( 50 mg ) . Tracer perfusion significantly increased brain d4-TCA levels 6 ± 0 . 1-fold with respect to vehicle treated controls ( n = 6–7 per group; p < 0 . 01 d4-TCA versus vehicle ) . These data strongly suggest that altered levels of circulating bile acids in the blood correspond to parallel changes in the central nervous system ( CNS ) . Although cocaine acts on centrally localized targets and TGR5 is expressed in the brain [2 , 17] , it is possible that TGR5 , FXR , or both of these receptors mediate the effect of OCA on cocaine behaviors . To discriminate between these possibilities , we first tested the involvement of TGR5 receptor signaling in the rewarding properties of cocaine by measuring cocaine CPP in TGR5 ( Gpbar1 ) knockout mice ( Gbpar1-/- ) . We found that deletion of the TGR5 receptor results in significantly increased preference for the cocaine-paired chamber relative to wild-type littermates ( Fig 3C ) . The enhancement of cocaine reward in the Gpbar1-/- mice identifies a role for the TGR5 receptor in reward processes and supports basal signaling through TGR5 as a contributor to resilience to cocaine reward . Furthermore , we showed that deletion of the TGR5 receptor precludes the effect of chronic OCA treatment on cocaine CPP ( Fig 3D ) , reinforcing the importance of TGR5 in mediating the effect of OCA treatment [2 , 17] . To localize the role of TGR5 in regulating cocaine behavior to the NAc , we utilized an adeno associated virus ( AAV ) vector to express GFP-tagged TGR5 or GFP in the NAc of Gpbar11-/- mice and measured cocaine CPP . Gpbar1 -/- mice virally expressing TGR5 in the NAc exhibited a significantly lower preference for the cocaine-paired chamber when compared to Gpbar1-/- mice virally expressing GFP ( Fig 3E ) . However , the question remains whether surgical GB-IL regulation of cocaine CPP requires TGR5 signaling . We performed GB-D and GB-IL surgeries in Gpbar1-/- mice and no differences were observed in cocaine CPP ( S5 Fig ) . These data reinforce the idea that the increase in circulating bile acids promoted by GB-IL requires TGR5 signaling in order to regulate the reinforcing properties of cocaine .
These findings support a role for bile acids and TGR5 signaling in neuronal function as well as in the control of motivated behaviors . This role was revealed by a novel surgery in which bile acids were diverted to the ileum to increase reabsorption and augment levels of circulating bile acids . We demonstrate that this surgery was able to modify reward acquisition and sensitization characteristic of chronic cocaine use . The GB-IL surgery blocks both sensitization and the rewarding properties of cocaine , which both rely on increases in extracellular DA levels . Notably , the surgery alters cocaine’s ability to increase DA levels in the NAc both in cocaine-naïve and cocaine-exposed animals . These results thus reveal that a surgery designed for weight loss also regulates psychostimulant reward . We found no alterations in gut microflora or striatal cocaine bioavailability in postoperative animals , making causative roles for microbiota to brain communication or reduced central cocaine concentration less likely . In order to exploit the utility of bile diversion surgery for translational opportunities , we sought to uncover the signaling pathways mediating the effect of the surgery on cocaine induced behaviors . The main direct effect of the surgery , elevating serum bile acids , led us to consider the receptors targeted by these circulating hormones as a likely candidate mediating our observed behavioral phenotype . Consistent with this hypothesis , we demonstrate that the ability of GB-IL to inhibit cocaine CPP requires TGR5 expression . Furthermore , we show that exogenous increases in bile acid signaling through OCA administration are sufficient to reproduce the effect of the GB-IL surgery on cocaine reward–related behavior in nonsurgical animals . While OCA more potently targets FXR bile acid receptors , our results point towards a dominant role of TGR5 in mediating the effect of elevated bile acid signaling on the behavioral response to cocaine . Specifically , Gpbar1-/- mice exhibit enhanced cocaine preference compared to their wild-type counterparts . Importantly , we show that OCA is incapable of altering cocaine conditioning in Gpbar1-/- animals . These data suggest that TGR5 is mediating the effects of OCA and that TGR5 represents a novel target for the modulation of motivated behaviors . However , this study does not fully address the pharmacology of OCA in the brain , and future studies are required to further define bile acid signaling in the NAc . Our results add novel gut signals ( bile acids ) as central regulators of drug reward–related behaviors . We present evidence that TGR5 may be acting within the NAc as viral re-expression of the receptor in the NAc of Gpbar1-/- mice reduces cocaine preference relative to GFP-Gpbar1-/- controls . The significance of TGR5 signaling is further supported by the inability of GB-IL surgery to reduce cocaine preference in Gpbar1-/- mice . Together , these results point to a role for central bile acid signaling in reducing susceptibility to cocaine and specifically implicate an accumbal receptor population . Future work could use more targeted manipulations to dissect the individual role of core and shell accumbal subregions in the effects described here . These results represent the first report of bile acids acting centrally to alter motivated behavior and open up novel avenues for translational investigations . Thus , further studies exploring whether pharmacologic , or even surgical , enhancement of bile acid signaling could intervene in models of established addictions are warranted . Importantly , the bile acid receptor agonist used in the current study ( OCA ) is on the market for the treatment of primary biliary cholangitis . This drug showed clinical efficacy in this setting with an excellent safety profile , thereby reducing barriers to its application for addiction treatment . Through the identification of the bile acid signaling system as an “already drugged” target to limit cocaine reward , this work delineates a significant advancement toward novel therapies for psychostimulant addiction .
Surgical analgesia is achieved with Ketoprophen ( 5–10 mg/kg , subcutaneous [s . c . [ , q 24 hr ) at the completion of the surgical procedure and additional supplementation is provided if required . For bariatric surgical procedures , analgesia coverage will be for 72 h and 48 h , respectively , and as needed thereafter . Following all experimental procedures , the animals are euthanized with sodium Pentobarbital ( 125 mg/kg , intravenous [i . v . ] ) , CO2 ( inhaled ) , or exsanguination under anesthesia . Veterinary care and oversight is provided by the School of Medicine’s Division of Animal Care , which is staffed with 4 veterinarians . An Ethics Committee within the Division of Animal Care approved the animal experiments of this study ( Protocol ID#: M/14/206 ) . Vanderbilt University is AAALAC accredited and operates under the principles outlined in the Guide for the Care and Use of Laboratory Animals ( DHEW Pub . No . ( NIH ) 86–23 Revised 1985 ) . Male wild-type C57BL/6J mice used for surgeries or for OCA treatment were acquired from Jackson Laboratories ( Bar Harbor , Maine ) at 5 weeks of age . Mice were acclimated to a Vanderbilt University housing facility for one week prior to surgery . Surgery ( GB-D or GB-IL ) occurred at 6 weeks of age . Mice were given at least 2 weeks to recover from surgery and were handled for 3 days prior to the start of the CPP paradigm . At this point , mice either underwent behavioral testing ( beginning with CPP ) or were sensitized to cocaine without behavioral testing . Gpbar1 ( TGR5 ) knockout heterozygous breeder mice were obtained from Dr . David Wasserman and generated as described in Vassileva and colleagues [22] . Heterozygous mice were mated to generate male and female knockout mice and wild-type mice used in behavioral experiments . The temperature- and humidity-controlled facility is maintained on a 12:12 h light:dark cycle ( lights on 07:00–19:00 h ) , and all experiments were performed during the light phase . The control surgery ( GB-D ) and experimental surgery ( GB-IL ) were performed as previously described [5] . Body weights were measured immediately prior to surgery and following surgery up until sacrifice and were averaged within 4 day bins . Following recovery from surgery , GB-D and GB-IL mice were treated with saline and cocaine at the dosing schedule used for cocaine CPP ( briefly , i . p . injections of saline every other day for 8 days and injections of cocaine on alternate days ) . Mice were euthanized 1–2 days following their final cocaine injection . Nucleus accumbens slices were prepared as previously described [23] . Mice were sacrificed by rapid decapitation under isoflurane anesthesia , and 300-μm slices were prepared with a vibratome in ice-cold oxygenated ( 95% O2/5% CO2 ) sucrose solution ( sucrose 210 mM; NaCl 20 mM; KCl 2 . 5 mM; MgCl2 1 mM; NaH2PO4·H2O 1 . 2 mM; NaHCO3 26 mM; dextrose 10 mM ) . Evoked DA release was measured in response to electrical stimulation using amperometry as described in Schmitz and colleagues [24] . Slices were maintained at 28 °C and continuously perfused with oxygenated artificial cerebrospinal fluid ( ACSF ) ( NaCl 125 mM , KCl 2 . 5 mM , NaH2PO4·H2O 1 . 2 mM , MgCl2 1 mM , CaCl2·2H2O 2 mM; NaHCO3 26 mM; dextrose 10 mM ) . Carbon fiber electrodes were fabricated by using a 7-μm carbon fiber ( Goodfellow , Coraopolis , PA ) and held at a voltage of +400 mV . Recordings were performed in the NAc core at a depth of 50–75 μm in response to a single electrical pulse ( 200 μA , 0 . 1 ms ) from a bipolar stimulating electrode . After stable control responses were established , 10 μM cocaine was applied to the slices . Mice were sacrificed by rapid decapitation under isoflurane anesthesia at 4–5 weeks following GB-D or GB-IL surgery . The brain was quickly dissected , blocked , and the NAc was punched bilaterally . Punches were stored at −80 °C until processing . To measure monoamines , high-performance liquid chromatography ( HPLC ) was performed as previously described . [25] CPP was performed as previously described , with modifications [26] . Briefly , 2-chamber CPP apparati ( MED-CPP2-MS; Med Associates , St . Albans , VT ) with distinct rod and mesh floor inserts were used . The associated software allowed for automated measurement of beam breaks on X–Y–Z axes ( 16 infrared beams , 50-ms intervals ) . Mice were weighed and then acclimated to the testing room for 20 min prior to testing each day . During the first phase ( pre-conditioning , day 1 ) , mice were placed on the grid floor side of the 2-chamber apparatus . For 30 min , the mice had free access to both sides of the apparatus . During the second phase ( conditioning , days 2–9 ) , on alternate days , mice were restricted to one side or the other of the apparatus for 30 min by use of a dividing door . Just prior to being placed in the chamber , each mouse was given an i . p . injection of either cocaine ( 20 mg/kg ) or saline . Cocaine was paired with the side of the apparatus less preferred during preconditioning . Approximately half of the mice were started on cocaine , while the other half were started on saline . During this time , each mouse’s locomotor activity was measured and used to determine cocaine-induced locomotor sensitization . The final phase of CPP ( post-conditioning , day 10 ) consisted of placing the mouse on the cocaine-paired side initially with the dividing door removed; however , no drug was given on this day . Thus , mice were given full access to both compartments and their time spent on each side was measured . Percent of CPP was calculated as the time spent on the cocaine-paired side during post-conditioning , minus the time spent on the cocaine-paired side during pre-conditioning divided by the time spent on the saline-paired side during pre-conditioning . The first 20 min of pre-conditioning and post-conditioning were used in the calculation of percent of CPP . All CPP was performed during the first phase of the light cycle . Activity Monitor v5 . 10 ( MED Associates ) was used to analyze CPP activity . Mice were administered cocaine ( 20 mg/kg , i . p ) and 30 minutes later , were sacrificed by rapid decapitation under isoflurane anesthesia . The brain was quickly dissected , blocked , and the striatum was punched bilaterally with a 0 . 75 mm inner diameter punch . Tissue punches were placed into Eppendorf tubes on dry ice and stored at -80°C until processing . Tissue was homogenized in 150 μL 100 mM sodium carbonate . Following centrifugation , samples were extracted using acetonitrile containing 50 nM cocaine-d3 as an internal standard . The resulting solution was dried under nitrogen and brought up in Mobile Phase A . Tissue cocaine content was quantified using LC-MS/MS on a Thermo TSQ Quantum ultra AM triple-quadrupole mass spectrometer in positive-ion mode using 0 . 1% HCOOH in Water ( solvent A ) and 0 . 1% HCOOH in Acetonitrile ( solvent B ) . The major transition , m/z 304 to 182 , was used to determine cocaine concentration . Four to seven days following CPP , GB-D and GB-IL mice from selected cohorts were tested for OF locomotion . Mice were initially weighed . Following 20 min of acclimation to the testing room , mice were placed in clean automated OF chambers ( 28x28 cm; MED-OFA-510; MED Associates ) under constant illumination for 60 min , and ambulatory distance was recorded . Mice were then removed from the chamber , injected with saline ( i . p . , equivalent to a 20 mg/kg dose of cocaine ) , and placed back in the chamber for 90 min . Finally , mice were removed again and injected with cocaine ( 20 mg/kg , i . p . ) before being placed back in the chamber for an additional 120 min . Following CPP and OF locomotion , GB-D and GB-IL mice from selected cohorts were tested on the HWM . The water maze protocol here was modified from a protocol previously described [27] . A round tub measuring 92 × 92 cm was filled with clean water the day before the first day of behavioral testing . On each morning of testing , mice were acclimated to the testing room for at least 10 min , after which behavioral testing began . For the first 5 days , a platform was placed just under the water in the northeast corner of the maze such that mice could not see it . Each day for 4 trials per day , mice were placed into the pool facing the wall and were given 60 seconds to find and stand on the platform . If they found it , they were allowed to stand on it for 10 seconds before being removed by the experimenter . If they did not find the platform in the 60 seconds given , they were placed on the platform by the experimenter for 20 seconds . After each trial , mice were allowed to dry in a clean cage on top of a warming pad , with at least 10 min in between each trial . On the final day of testing , the platform was removed . The mice were placed in the pool for a single trial and percentage of time in the target quadrant was measured . Following CPP and OF locomotion , GB-D and GB-IL mice from selected cohorts were tested on the TST . This involved individually suspending each mouse by the tail using adhesive tape to a flat , stainless steel force sensor connected to a computerized monitoring system ( v3 . 30 , MED Associates ) . The force sensor measured the amount of time each mouse spent struggling to right itself . The mouse was suspended from the sensor for a total of 6 min . The last 4 min of the trial were used to calculate time immobile , which was defined as the total time during which the mouse movement did not exceed a preset threshold of seven for 200 ms . Following CPP and OF locomotion , GB-D and GB-IL mice from selected cohorts were tested on the rotarod . The rotarod consisted of a rotating , grooved rubber cylinder ( approximately 3 cm in diameter ) . Mice were placed on the cylinder , which rotated for 5 min , gradually increasing from 4 to 40 rpm . The amount of time spent on the cylinder before safely falling was recorded . Serum bile acids were measured by mass spectrometry using methods previously described [5] . Bile acids were measured from trunk blood taken immediately following decapitation at sacrifice at 4–5 weeks and 7–8 weeks post-surgery . To allow for gut bioavailability of the semi-synthetic bile acid analogue OCA ( chemically 6-ECDCA or 6α-ethyl-chenodeoxycholic acid , AdipoGen , San Diego , CA , #AG-CR1-3560-M025 ) without the stress of oral gavage , OCA was administered to mice by voluntary oral administration . OCA was initially dissolved in beta cyclodextrin ( 20% w/v ) and then dissolved within palatable drug-laced jellies . Jellies were composed of gelatin ( 10% w/v ) , sucralose ( 18 . 5% w/v ) , artificial strawberry flavoring ( 8% v/v ) ; beta cyclodextrin ( 2% w/v ) in water . Jellies containing OCA or beta cyclodextrin without dissolved drug were made to contain 10 mg/kg based on each mouse’s original weight on the first day of drug or vehicle administration . They were given the jellies by placing each mouse into an OF chamber containing the jelly for 20–30 min on six consecutive days per week for 4 weeks . To ensure that mice consumed the jellies consistently , all mice were initially trained to eat jellies without the drug for 5 days prior to drug/vehicle jelly administration . Cecal content samples were collected from GB-D and GB-IL mice at sacrifice 4–5 weeks after surgery and stored at −80 °C . Microbiota analysis was performed as previously described [5] . Both the AAV2/5-Gpbar1 ( AAV2/5-CBh-m-Gpbar1-T2A-eGFP-WPRE , 3 . 6 × 1012 GC/ml ) and GFP ( AAV5-GFP , 1 . 0 × 1013 GC/ml ) vector were obtained from Vector BioLabs ( Philadelphia , PA ) . The AAV5-GFP solution was diluted in 5% glycerol PBS to match GC content of AAV2/5-Gpbar1 . At 8 weeks of age , mice were anesthetized via isoflurane inhalation and given 0 . 5-μl bilateral microinjections of AAV2/5-Gpbar1 or AAV5-GFP at a rate of 1 nL per second into the NAc ( A/P + 1 . 3 mm , M/L +/− 1 . 0 mm , D/L −4 . 0 mm , measured from Bregma ) [28] . Behavioral assays occurred three weeks after viral injection in order to coincide with the peak of AAV-mediated transgene expression [29 , 30] . Following behavioral assays , animals were sacrificed and viral expression was examined . We excluded animals that did not exhibit consistent bilateral viral expression throughout the NAc . To confirm appropriate transduction and targeting of viral injections , mice were perfused with 4% paraformaldehyde in PBS and the intact brains were removed , postfixed for 24 hours , cryoprotected with 20% sucrose ( PBS ) overnight , and then sectioned and processed . Brains were sectioned ( 50 μm ) using a Leica VT1000s ( Buffalo Grove , IL ) and stored in 0 . 1 M phosphate buffer . To stain , slices were permeabilized in 0 . 1% Triton X-100 ( Thermo Fisher Scientific , Waltham , MA ) in 2% goat serum ( Jackson ImmunoResearch , West Grove , PA ) . Endogenous peroxidases were quenched with 1 . 0% sodium borohydride and 0 . 15% hydrogen peroxide ( Sigma , St . Louis , MO ) . Slices were incubated in rabbit anti-GFP ( Abcam ab290 1:2 , 500 at 4 °C , overnight ) and HRP-conjugated goat anti-rabbit secondary antibody ( Santa Cruz , CA , 1:200 at room temperature for two hours ) . Signal was amplified using a TSA Cyanine 3 system ( Perkin Elmer , Waltham , PA ) . Data are presented as means ± standard error of the mean . Statistical analysis was performed with GraphPad Prism software , version 6 ( GraphPad Software , San Diego , CA ) using statistical tests noted in figure legends . Outliers were defined as having values outside of quartile 1–1 . 5 × interquartile range ( IQR ) and quartile3 + 1 . 5 × IQR and were excluded . A p value < 0 . 05 defined statistical significance for all tests . | Communication between the gut and the brain is increasingly being appreciated as influencing motivated behavior . The gut can influence brain function through secreted hormones traveling through the blood and entering the brain . We utilize a weight-loss surgery designed to elevate one class of circulating hormones , bile acids , to show their action in the brain and their role in modulating behaviors associated with the addictive properties of cocaine . This surgery reduces the reward-related behavior and the psychomotor effects of cocaine . Furthermore , we utilize a knockout mouse model to reveal that a specific bile acid receptor mediates some of the effects of bile acids over motivated behavior . Viral intervention studies localize this effect to a receptor population within the nucleus accumbens , a brain region central to the processing of reward . These findings identify a role for bile acids in blunting cocaine’s ability to alter brain function , generating novel and exciting directions for the treatment of cocaine abuse . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biliary",
"system",
"alkaloids",
"medicine",
"and",
"health",
"sciences",
"liver",
"body",
"fluids",
"chemical",
"compounds",
"gallbladder",
"bile",
"vertebrates",
"mice",
"animals",
"mammals",
"biological",
"locomotion",
"surgical",
"and",
"invasive",
"medical",
"pr... | 2018 | Bile diversion, a bariatric surgery, and bile acid signaling reduce central cocaine reward |
The -function and the -function are phenomenological models that are widely used in the context of timing interceptive actions and collision avoidance , respectively . Both models were previously considered to be unrelated to each other: is a decreasing function that provides an estimation of time-to-contact ( ttc ) in the early phase of an object approach; in contrast , has a maximum before ttc . Furthermore , it is not clear how both functions could be implemented at the neuronal level in a biophysically plausible fashion . Here we propose a new framework – the corrected modified Tau function – capable of predicting both -type ( “” ) and -type ( “” ) responses . The outstanding property of our new framework is its resilience to noise . We show that can be derived from a firing rate equation , and , as , serves to describe the response curves of collision sensitive neurons . Furthermore , we show that predicts the psychophysical performance of subjects determining ttc . Our new framework is thus validated successfully against published and novel experimental data . Within the framework , links between -type and -type neurons are established . Therefore , it could possibly serve as a model for explaining the co-occurrence of such neurons in the brain .
Monocular presentation of a looming object elicits escape or avoidance reactions in many species , including humans [1]–[4] . When a planar object travels perpendicular to a surface toward an observer ( i . e . the object approaches the observer on a direct collision course ) , it projects a symmetrically expanding image on the retina . Notice that in the present paper we only focus on a subset of approaches where the approaching object eventually collides with the observer . We assume that collision happens at time ( time to contact , “ttc” ) . At time before , the image subtends an angle , and its outer contours expand with angular velocity . Both angular variables grow nearly exponentially with decreasing distance between object and eye ( assuming a constant velocity ) . With knowledge of a predator's or object's typical size [5] , it is therefore possible to trigger a behavioral response as soon as or , respectively , crosses a threshold [1] , [6] , [7] . The visual systems of various species are also known to “compute” functions of and ( see e . g . [8] for a recent review ) . The Tau-function ( “” ) is defined by . Under the assumption that the object is a rigid sphere that approaches with , has several interesting properties [9] , [10]: First , provides a running estimation of ttc during the approach . Second , the ttc estimation is largely independent of physical object size , provided that and are noise-free . Third , decreases approximately linearly with time with a constant slope of , but eventually linearity is compromised , as has a minimum shortly before ttc . It therefore would be comparatively easy to track the remaining time until impact , and to precisely time avoidance reactions , for example as soon as is below a certain threshold value . These three properties , however , are valid only for “sufficiently small” angular sizes . Any quantitative criterion for “sufficiently small” implicates an error threshold for the deviation of from linearity , that is . For example , according to Text S6 a corresponding threshold for the visual angle can be defined as with some constant . Notice that the -criterion is independent from stimulus parameters such as object diameter or approach velocity . Because is well suited for the estimation of , it could in principal serve as a universal mechanism for guiding motor actions during object approaches or during self-motion towards static objects . Indeed , several studies related to behavioral responses in this context , thus asserting that many organisms , including humans , rely on for their timing of motor actions ( e . g . [10]–[12] ) . But a critical re-evaluation of the -hypothesis arrived at the conclusion that does not necessarily play a unique role for ttc estimation [13] , [14] . For example , humans also rely on the rate of change of relative disparity , particularly in the late phase of an approach , for small object sizes [15]–[18] , for low speeds [19] , [20] , or if knowledge of object size is available [7] . In addition , the task at hand ( e . g . catching a ball or eluding a meteorite ) seems to dictate the information that will eventually be used for action timing [14] , [18] , [21] . Further inconsistencies with respect to were reported with psychophysical results , where tended to be underestimated [16] . In addition , ttc -estimation reveals a certain dependence on object size [22] , which is also not predicted by at “sufficiently small” angular sizes . The Tau-function is often studied in the context of ttc -estimation . It appears , however , that in order to describe the responses of collision-sensitive neurons in certain species is inadequate . For example , the Lobula Giant Movement Detector ( LGMD ) neuron in locusts responds with increasing activity to a stimulation with a symmetrically expanding image , if the expansion pattern is consistent with an approaching object [23] , [24] . The response curve of the LGMD neuron gradually increases to a maximum and then abruptly ceases ( often to a nonzero baseline response ) . Because does not have a maximum , a different function has been proposed for modeling LGMD responses: The Eta-function ( “” ) . It is defined as , with a constant [25] . Theoretically , the time when the activity peak occurs depends linearly on the ratio of object half-size to object velocity . The peak will shift closer to for smaller or faster objects , and always occurs at angular size , independent of [26] . The LGMD activity peak could in principle signal a critical angular size for escaping . Indeed , a recent study with freely behaving locust suggests that the time of peak firing rate of the Descending Contralateral Movement Detector ( DCMD ) predicts that of jump [27] ( each LGMD spike triggers a spike in the postsynaptic DCMD as well , because the LGMD is strongly coupled to the DCMD by a combined electrical and chemical synapse [28] , [29] ) . It has nevertheless been argued that – in some ecologically meaningful situations ( small ) – there is no guarantee for the peak to occur before [2] , [5] . This statement may be true to the extent that in freely behaving locusts , a reliable escape jump is triggered before collision only in the range of to [30] . For , the jump would occur after projected collision , and this value thus may reflect the typical sizes and speeds of predators . Apart from the locust , other species have collision-sensitive neurons with -like properties , for instance fruitflies [31] and bullfrogs [32] . In pigeons , the response properties of one of three classes of neurons in the dorsal posterior zone of the nucleus rotundus also seems to be compatible with the -function [1] . ( The two remaining classes seem to compute and , respectively ) . In the goldfish , responses of the M-cell to looming stimuli also appear to follow a version of the -function , in which replaces , such that the new function does only depend on [33] . The Tau-function and the Eta-function are the two prevailing models for studying ttc -perception and ( interceptive ) action timing on the one hand , and escape behavior and collision avoidance on the other . In other words , we have two different models for two seemingly separated contexts . Each model brings about some hitherto unresolved issues , which are subsequently described . From a computational point of view , is numerically unstable: In the presence of noise , we have to reckon with the fact that can get very small – or even reach zero – at certain instants during the initial phase of the approach ( cf . [17] ) . As a consequence , fluctuations of with large amplitudes may occur . If , however , noise levels are constant in time , and noise is not multiplicative , the signal to noise ratio continuously improves as is approached . It is furthermore not entirely clear how could be biophysically implemented in a neuron . As for the -function , the LGMD neuron seems to bypass a direct multiplication or division by computing with subsequent exponentiation of the result [34] . From a mathematical viewpoint , however , taking the logarithm introduces an instability for , although neuronal circuits with divisive inhibition can be adjusted such that no stability problems occur [35] . Moreover , Gabbiani et al . [34] found that a third-order power law fitted the mean instantaneous firing rate of the LGMD better than an exponential or a linear function ( see also reference [36] ) . Our original motivation was to improve the stability of with a simple modification . This modification led us to the modified Tau function . Similar to the -function , the -function also reveals a maximum before ttc . We were able to fit the response curves of -type neurons with ( Text S4 ) . Our -function represents the equilibrium solution of an equation for describing neuronal firing rate . Because of this , is based on a biophysically plausible mechanism . But comes with a disadvantage: Unlike , it no longer provides a running value of ttc . In order to recover the ttc prediction , we needed to add a correction term to . This so-defined corrected modified Tau function ( ) recovers the ttc prediction of the original -function , but suppresses noise better than . Most importantly , the corrected m-Tau function predicts the results of a psychophysical experiment , requiring subjects to estimate ttc . Theoretically , we therefore can explain -type and -type responses within the framework , which contains ( but also ! ) as a special case . Until now , and did not have any obvious relationship with each other ( although we show in Text S6 how could formally be related to ) . The -function could thus serve to explain why -type and -type neurons could be found alongside each other in the pigeon brain [1] .
Behavioral and neural responses to optical variables ( e . g . , , , , ) in the initial part of a trajectory are very noisy signals . Signal fluctuations may occur as a consequence of the discrete structure of the retinal photoreceptor array and its limited spatial resolution . The signal-to-noise ratio continuously improves as ttc is approached ( Text S3 ) . Our first step adds computational stability to the model . Let be a constant ( in units of ) . The modified Tau model is defined as: ( 1 ) Biophysically , can be interpreted as leakage conductance ( equation S2 in Text S1 ) . According to equation ( 1 ) , can formally be expressed in terms of multiplied with a gain control factor , which depends only on angular velocity . Notice , however , that the multiplicative version “” would again compromise stability , because appears as one of the factors in the product . Figure 1 a juxtaposes and the factors and , respectively . Let the initial distance between the eye and a circular object ( diameter ) be denoted by . Then , choosing will create a maximum of at time ( i . e . , a maximum before ) : ( 2 ) ( the previous equation is derived in the Methods Section ) . The time when assumes its maximum can thus be controlled by specifying , where bigger values will place the maximum closer to . The maximum depends as follows on approach velocity and object diameter , respectively . Assume fixed values for and . Then , will have an activity maximum at ( default case ) . Now increase approach velocity and initial distance , such that remains constant . As a consequence , the peak will shift closer to with respect to the default case ( triangle symbols in Figure 1a; further figures in Text S2 ) . This is the velocity effect . Now increase the object diameter . The maximum of will then occur earlier compared to the default case ( circle symbols in Figure 1 ) . This is the size effect . Assuming that the peak signals an imminent collision , this shifting behavior is consistent with larger objects being perceived to have an earlier ttc than smaller ones [22] . Note that the original -function ( i . e . and noise-free angular variables ) does not show a strong dependence on object size where holds ( but see Text S6 ) . The -function is the prevailing model for describing responses from collision sensitive neurons to object approaches with constant velocity . Its characteristic feature is its maximum . Because also has a maximum , we fit previously published neuronal response curves with the -function and ( Text S4 ) . Figure 2 summarizes these fits by comparing the response maxima of the experimental curves ( “” ) with the maxima predicted by the best fits achieved with the two functions ( “” ) . Predictions of are slightly better with -fits , both in terms of mean and median of absolute differences ( ) . With respect to goodness of fit measures ( root-mean-square-errors , , F-statistics ) , both functions perform again on par with each other . Therefore , both and the -function describe neuronal responses of object approaches with constant velocity . The experimental maxima at time depend linearly on [26] . The -function predicts this linear relationship ( equation S5 in Text S2 ) , where slope is identified by , and intercept by a temporal delay of corresponding line fits ( Figure 3a ) . The maximum of the -function , however , depends in a nonlinear way on ( equation 2 & equation S6 in Text S2; illustration: Figure 4 ) . ( Nonlinearity means that the slope depends on , and linearity means that it does not ) . Linearity is approached with increasing values of , eventually reaching a slope of one for ( equation S9 in Text S2 ) . This is nevertheless inconsistent with experimental evidence , as the experimental values for are underestimated ( typically ) . We thus explored a different possibility: Can the nonlinear function be hidden by noise ? Figure 3b suggests that it nearly can , as seen when fitting a line to a version of with additive Gaussian noise . Noise levels were set as reported in [26] . This hide-and-seek works quite well , and the fitting statistics ( , KS-test on residuals , F-statistics ) are consistent with linearity in many random trials ( detailed analysis: Text S2 ) . Figure 4 suggests a correlation between intercept and slope of line fits for different values of . We thus fit lines to the noisified version of for various values of . As before , noise levels were set as reported , and we again identified intercept and slope of the line fits to with and , respectively . The result of this procedure is shown in Figure 5 , and agrees well with Figure 4 in [26] . Thus , consistently predicts a good correlation between intercepts and slopes both in the presence and in the absence of noise . Maximum detection of in the initial phase of an object approach ( i . e . , for small values of ) is problematic , due to the signal's poor signal-to-noise ratio and the rather “shallow” curvature around the maximum . The situation gets progressively better if we place the maximum closer to , that is for bigger values of : The signal-to-noise ratio is better , and curvature is higher . With , however , we fell short of explaining the results of our psychophysical experiment ( which is below described further ) . This led us to modify as follows . Observe that for all , and thus ( 3 ) is a positive correction factor to , such that . As with , the correction factor per se is again susceptible to fluctuations in the angular variable , and we would have gained no improvement by simply adding it to . Now , the crucial idea is to render insensitive to such fluctuations . This is achieved with a first order low-pass filter ( a short introduction is given in Text S8 ) . Low-pass filtering of and transforms into a slowly varying signal , which is eventually added to : ( 4 ) and are low-pass filtered visual angle and angular velocity , respectively , and is the system's integration time constant . In order to avoid initial filter transients , the filter variables were initialized with and , respectively . The are filter memory coefficients with for . No filtering would take place for ( no memory ) , and the filters would never change their initial state for ( infinite memory ) . The corrected , modified model ( “corrected m-Tau” ) is then defined as: ( 5 ) where is a small constant , such that possible division-by-zero errors are avoided in the simulation . Nevertheless , in the presence of noise , division-by-zero errors do not typically represent a problem during an approach with , because if the following two conditions hold: ( i ) appropriate initialization of , and ( ii ) “sufficiently strong” lowpass filtering . The offset is included for the sake of completeness . It was only considered for simulating our psychophysical experiment ( described below ) , where it turned out to be negligibly small . In general , therefore , it is safe to assume . Similar to , the new corrected m-Tau-model also computes an estimation of ttc for “sufficiently small” angular sizes . But the principal advantage of over is that it is less sensitive to noise . The noise suppression of the corrected m-Tau-model is constrained by the noise suppression performance of two “limit functions” , which are approached dependent on the values of , , and ( Figure 6 ) . For the derivation of these limit functions , assume ( to simplify matters ) that in equation ( 5 ) with ( and ) . Then , as we will show subsequently , the constraining functions are the ordinary function for , on the one hand ( equation 6 ) , and for a version of with lowpass-filtered angular variables , on the other ( equation 8 ) . Thus , , where , provided that we exclude the case , , which would imply that is unbounded . Details on our psychophysical experiment are spelled out in the Methods Section . In a nutshell , subjects viewed approaching balls on a monitor . The balls had two different sizes ( big & small , corresponding to object diameters & , respectively ) , and disappeared after ( presentation time ) until . A beep sounded always at the same time , , in order to indicate a reference time to the subjects . Approaches with different values of were presented , where could occur before or after . Subjects were asked to judge whether they were hit by the ball before or after . Responses were pooled , and the “proportion of later responses” for each presentation time ( corresponding to “ball hit me after ” ) was computed as a function of ttc . Figure 7a shows the corresponding data points for , along with the best matching Gaussian cumulative density function ( “GCDF”-fit ) for each object diameter . The GCDF-fits represent an estimate of the underlying psychometric curves or psychometric functions , respectively . Figure 7b suggests that subjects did not respond to the average of the stimulus set , because the mean of the distribution ( point of subjective simultaneity ) shifted with presentation time . In addition , the variance of the distribution decreased with increasing presentation time . The small object diameter is furthermore associated with a higher variance than the big one . The full set of data points is shown in Figure 8 , where each figure panel corresponds to a different presentation time ( small object size: circles; big: triangles ) . The curves shown in Figure 8 do not represent GCDF-fits ( as in Figure 7a ) , but rather display simulation results from the -model . For short presentation times , subjects show near-random performance across ttc ( Figure 8a , b ) , thereby revealing a bias towards later responses ( i . e . “ball hit me after ” ) . The GCDF-fits reveal a higher bias for the small object diameter ( Figure 7b ) . The corresponding psychometric functions ( not shown ) and -predictions for the shortest presentation time ( ; Figure 8a ) are thus rather flat and noisy . This bias is progressively reduced with increasing , indicating improving performance: For , the point of subjective simultaneity approaches for both object diameters , and psychometric functions get closer to a step-wise increase at ( Figure 7a ) . We already mentioned that we simulated the psychometric functions with the corrected m-Tau -model ( equation 5 ) , at which we added noise to angular size and angular velocity ( equation 9 ) . By assuming a constant approach velocity , one could compute an estimation of ttc with equation ( 12 ) . Note that this estimation should be constant throughout the approach in a noise-free situation and for “sufficiently small” angular sizes . As a consequence of having noise , however , the ttc estimation fluctuates . We therefore computed an average estimation with equation ( 14 ) , by taking the mean value across a time interval ( the interval contained the last estimates ) . The average ttc estimation was evaluated at presentation time , and compared with the reference . With a total number of such trials , we then counted occurrences where the average estimate occurred after . The simulated proportion of later responses is then obtained by dividing by ( equation 13 ) . In order to find the appropriate -parameters for predicting psychophysical performance , the error between -predictions and psychophysical data points was minimized . We refer to this procedure as optimization . Optimization was carried out separately for object diameters big and small . The first step of the optimization procedure consisted in parsing the parameter space , and recording the error associated with each set of -parameters . The error was determined with two measures ( “score measures” ) : The root mean square error ( ) , and an outlier-insensitive robust error ( ) . In the second step , the parameter sets were sorted in ascending order with respect to their associated score measure . Sorting took place separately for and , leading to corresponding tables where the best set of parameters was assigned rank one ( 1st table row ) , the second best rank two ( 2nd table row ) , and so on ( Tables S1 & S2 in Text S5 ) . A third table of -parameters was then computed which was optimal for both object diameters simultaneously ( combined; Table S3 in Text S5 ) . This could be done in a straightforward way , simply by averaging the score measures of big and small of corresponding parameter sets , and subsequently sorting the averaged errors ( more details on finding parameters are given in Text S5 ) . For the computation of and , all psychophysical data points that represent the proportion of later responses entered equivalently , in the sense that no weighting coefficients were used to bias the optimization process toward longer presentation times ( as GCDF-fits at longer presentation times have a smaller variance , see Figure 7b ) . Notice that parameter optimization for the combined diameter naturally implicates a trade off – the errors with respect to big and small will be bigger compared to individual parameter optimization . Figure 8 shows that the corrected m-Tau -model adjusts fairly well to the psychophysical data of both object diameters . Nevertheless , the -predictions for are somewhat worse with the combined parameter optimization ( Figure 8e ) when compared to a separate optimization for big and small ( corresponding figures in Text S7 ) . The most likely explanation for this discrepancy ( individual versus combined parametrizations ) is that each object size is associated with a different noise level ( noise levels are represented by the -parameters with ; see equation 9 ) . We investigated this hypothesis by comparing the corresponding values of for big and small , as a function of their rank . Figure 9 shows that the for small are consistently higher than for big . Therefore , the corrected m-Tau -model generally supports the notion that smaller object diameters imply higher noise levels in angular size and angular velocity , respectively . We also studied two models with less degrees of freedom than corrected m-Tau : The first was , and the second was with for ( ) . The best ( i . e . smallest ) score measures achieved with these reduced models were consistently higher than the best values achieved by the corrected m-Tau -model ( Text S5 ) , and their best-ranked parameter sets resulted in psychometric curve predictions that were also inferior by visual inspection ( not shown ) .
With the corrected m-Tau -model equation ( 5 ) , we proposed a general framework that comprises the -function and several properties of the -function . By means of adjusting only a single parameter ( ) , the corrected m-Tau -model can approximate and , respectively . Moreover , the -approximation is less sensitive to noise than the original -function , and accounts well for the performance of the psychophysical experiment that we carried out . In the experiment , subjects had to decide whether a ( displayed ) ball reached them before or after a reference signal at time . However , balls were only presented until , and disappeared afterwards . In other words , subjects had to estimate ( could occur before or after ) . With respect to our experiment , the corrected m-Tau -model suggests the following conclusions: The modified -model ( “” ) constitutes a special case of . It is obtained from equation ( 5 ) for ( by default ) . Its distinguishing feature is a maximum before , which can be shifted via ( equation 2 ) . The -maximum decreases as it is positioned closer to , because this implies bigger values of . The time of the -maximum depends furthermore on size and velocity ( Figure 1 ) . The curve shape of is reminiscent of the -function , since both functions have a maximum . We thus decided to fit previously published response curves from collision sensitive neurons to both functions , and observed that both functions fit the neural curves well in terms of goodness-of-fit criteria ( Text S4 ) . We must not forget , however , two important differences between and . First , since reveals a minimum shortly before ( Text S6 ) and derives from , the -response is more precisely biphasic . The biphasic structure gets pronounced in some of the curve fits , especially when is close to ( see corresponding figures in Text S4 ) . Then , the amplitude of the -maximum is small , and consequently the fitting algorithm has to scale it to the maximum of the neuronal recording data . In this way , the minimum is also scaled . Second , depends in a nonlinear way on the size-to-velocity ratio ( see Figure 4 for an illustration ) . This is contradictory to several studies that found a linear dependence . A linear dependence is also predicted by the -function ( equation S5 in Text S2 ) . The contradiction can be alleviated by adding noise to relative time of the -maximum ( ; equation S10 in Text S2 ) , with noise amplitudes as reported in [26] . As a consequence of noise , the nonlinear relationship can be literally hidden ( Figure 3 ) , such that statistical tests would affirm an underlying linear process ( Text S2 ) . Masking by noise is more effective for bigger values of , because the noise level is proportional to . The -function in its original form cannot explain the neuronal response curves for an approach with ( “linear approach” ) [25]: Rather than predicting a decreasing response with time , the -function would linearly increase . In contrast , the -function makes correct predictions . Correct predictions with can nevertheless be made by including an additional inhibitory process in the firing rate equation of ( equation S3 in Text S1 , where a full proof of concept is described ) . Important , this extension of ( i ) is based on a power function with an exponent between and , but not on an exponential function as with , and in this regard it may hence be considered as being biophysically more plausible than ( see also reference [36] ) ; ( ii ) does not interfere with the “normal” behavior ( i . e . normal object approaches are not affected ) ; and ( iii ) tolerates high noise levels ( i . e . , the mechanism is robust ) . What about alternative models which also have a response peak ? In Text S6 we studied two such functions , namely “inverse ” ( ) , and angular acceleration ( ) . Both of them reveal a linear dependence of on ( equations S24 & S26 , respectively , in Text S6 ) . The maximum of always precedes that of . However , does not make correct predictions for the “linear approach” , as we would obtain ab initio for ( although a dynamical version may predict the decreasing LGMD-activity on the basis of temporal filtering effects ) . In contrast , would make consistent predictions in that case . Without further modifications , though , neither nor seems to be adequate for fitting the response curves of collision sensitive neurons , because there is no free model parameter to shift their respective maximum . Although the occurrence of their maxima could principally be controlled by a global shift of the time scale , the corresponding values ( obtained by fitting the neuronal response curves ) would overestimate experimental values ( Text S6 ) . Similarly , when “fitting” the -function to and the so obtained values of would underestimate experimental values: The -maximum would coincide with the maximum of for , and with the maximum of for . In conclusion , is no replacement for the -function , at least for describing neuronal responses of collision sensitive neurons in insects . However , in the nucleus rotundus of pigeons three classes of neurons were reported [1] , [38] . They conform to -like , -like , and -like responses . The fact that is just a special case of could possibly explain why neurons with -like and -like properties can be found in a single brain . Within the -framework , the function corresponds to , and is obtained for choosing . Thus , the adjustment of only a single weight ( ) is necessary to go from one function to the other . The corrected m-Tau -framework could thus offer a parsimonious yet full-fledged explanation of the implementation of -like and -like neurons at the circuit level .
We simulated our psychophysical experiment with the corrected m-Tau -model ( equation 3 ) , where we plugged in noisified versions of the optical variables ( i . e . ) , ( 9 ) with noise probabilities ( ) , and with the dot denoting the time derivative . The are random variables , which at each instant return a value that is drawn from a centered normal distribution . In the last equations , we used the explicit expression for angular size , ( 10 ) and angular velocity ( rate of expansion ) ( 11 ) with and . The values of and are the psychophysical stimulus parameters . Simulations were carried out with a temporal resolution of . The corrected m-Tau -model is constrained by two limit functions: Ordinary on the one hand ( equation 6 ) , and on the other ( equation 8 ) . Both limit functions decrease approximately as ( illustration: Figure 6 ) . Thus , a ttc estimation at time can be computed as ( 12 ) ( Nomenclature: is the model prediction for ttc at time , and is the experimentally set parameter ) . In the psychophysical study , subjects were asked to estimate whether they were hit by the approaching object before or after . We accordingly define their proportion of later responses as the number of trials ( where subjects responded with being struck after ) divided by the total number of trials : ( 13 ) is represented by circle and triangle symbols in Figure 7 and 8 . The corresponding predictions from the model are denoted by . Specifically , with and , and analogous for . Computation of is required for , which we did with equation ( 12 ) as per ( 14 ) Notice that , due to noise ( equation 9 ) , will be subjected to random jitter with each trial . Therefore , in order to obtain a more robust estimate of ttc , we do not use only : The integral in the last equation computes – in the discrete case – the mean value across the last time steps until ( typically , what amounts to a time interval for averaging of , cf . first figure in Text S7 ) . In order to illustrate the noise level at each , we also computed the standard deviation of the last values of . The shaded areas in the figures which visualize & correspond to . Predictions of the corrected m-Tau -model are shown as curves in Figure 8 , as well as in Text S7 . The corrected m-Tau -model has eight free parameters: , , , , , , , . The parameter space was parsed with constant step widths . For each set of parameter values , -predictions for the proportion-of-later-response curves were computed according to the procedure described in the previous section . The corresponding “goodness of prediction” ( or “prediction performance” ) was evaluated with the root mean square error ( rmse , ) , and the outlier insensitive , robust error ( robe , ) , see equation S18 in Text S5 . The “goodness of prediction” measures are referred to as score-measures ( rmse-scores & robe-scores , respectively ) . Parameter values were sorted according to their scores . In this way we ended up with several score tables , which list the best set of parameters , according to object size: Table S1 in Text S5 for small object diameter ( ) , Table S2 in Text S5 for big object diameter ( ) , & Table S3 in Text S5 for combined object diameter . The scores for the combined size were computed by averaging the scores of big & small for corresponding parameter values , and then sorting the averaged scores in ascending order . More details on parameter finding and analysis are given in Text S5 . Consider a rigid sphere ( object radius or half-size ) that approaches an observer on a direct collision course . If the approach proceeds at a constant velocity , the object-observer distance at time is . Thus , the initial distance is . Now , consider the gain control factor from equation ( 1 ) ( 15 ) where we plug in the explicit expression for angular velocity equation ( 11 ) and obtain ( 16 ) Especially in the initial phase of the approach , when visual angle and angular velocity are sufficiently small , decreases approximately linearly with time ( cf . Text S6 ) , ( 17 ) Because of , the m-Tau function becomes approximately ( 18 ) A maximum of the m-Tau function implies that its first time derivative is zero . We define . The first time derivative of the ( approximate ) m-Tau function ( 19 ) disappears if , or ( 20 ) The last equation is the distance ( positive sign ) where the approximated m-Tau function attains its maximum during an object approach . Thus , the time when the -maximum occurs is ( 21 ) | In 1957 , Sir Fred Hoyle published a science fiction novel in which he described humanity's encounter with an extraterrestrial life form . It came in the shape of a huge black cloud which approached the earth . Hoyle proposed a formula ( “” ) for computing the remaining time until contact ( “ttc” ) of the cloud with the earth . Nowadays in real science , serves as a model for ttc -perception for animals and humans , although it is not entirely undisputed . For instance , seems to be incompatible with a collision-sensitive neuron in locusts ( the Lobula Giant Movement Detector or LGMD neuron ) . LGMD neurons are instead better described by the -function , which differs from . Here we propose a generic model ( “” ) that contains and as special cases . The validity of the model was confirmed with a psychophysical experiment . Also , we fitted many published response curves of LGMD neurons with our new model and with the -function . Both models fit these response curves well , and we thus can conclude that and possibly result from a generic neuronal circuit template such as it is described by . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"signal",
"processing",
"neuroscience",
"signal",
"filtering",
"cognitive",
"neuroscience",
"behavioral",
"neuroscience",
"mathematics",
"computational",
"neuroscience",
"circuit",
"models",
"biology",
"differential",
"equations",
"visual",
"system",
"calculus",
"psychophysic... | 2012 | Unifying Time to Contact Estimation and Collision Avoidance across Species |
The small GTPase Rab27a has been shown to control membrane trafficking and microvesicle transport pathways , in particular the secretion of exosomes . In the liver , high expression of Rab27a correlates with the development of hepatocellular carcinoma . We discovered that low abundance of Rab27a resulted in decreased hepatitis C virus ( HCV ) RNA and protein abundances in virus-infected cells . Curiously , both cell-associated and extracellular virus yield decreased in Rab27a depleted cells , suggesting that reduced exosome secretion did not cause the observed effect . Instead , Rab27a enhanced viral RNA replication by a mechanism that involves the liver-specific microRNA miR-122 . Rab27a surrounded lipid droplets and was enriched in membrane fractions that harbor viral replication proteins , suggesting a supporting role for Rab27a in viral gene expression . Curiously , Rab27a depletion decreased the abundance of miR-122 , whereas overexpression of miR-122 in Rab27a-depleted cells rescued HCV RNA abundance . Because intracellular HCV RNA abundance is enhanced by the binding of two miR-122 molecules to the extreme 5’ end of the HCV RNA genome , the diminished amounts of miR-122 in Rab27a-depleted cells could have caused destabilization of HCV RNA . However , the abundance of HCV RNA carrying mutations on both miR-122-binding sites and whose stability was supported by ectopically expressed miR-122 mimetics with compensatory mutations also decreased in Rab27a-depleted cells . This result indicates that the effect of Rab27a depletion on HCV RNA abundance does not depend on the formation of 5’ terminal HCV/miR-122 RNA complexes , but that miR-122 has a Rab27a-dependent function in the HCV lifecycle , likely the downregulation of a cellular inhibitor of HCV gene expression . These findings suggest that the absence of miR-122 results in a vulnerability not only to exoribonucleases that attack the viral genome , but also to upregulation of one more cellular factor that inhibit viral gene expression .
Hepatitis C virus ( HCV ) is a hepatotropic positive-sense , single-stranded RNA virus that belongs to the Flaviviridae family . The HCV genome is about 9 . 6 kb in length and encodes a polyprotein , which is cleaved into at least ten viral proteins by host and viral proteinases [1 , 2] . The open reading frame is flanked by 5’ and 3’ noncoding regions , which regulate translation and replication of the viral RNA . In addition , the 5’ terminal sequences of the HCV RNA genome form an oligomeric complex with two molecules of liver-specific miR-122 [3 , 4] . This complex greatly stabilizes the viral RNA from degradation by exonucleases [5 , 6] . Exposure to HCV typically leads to persistent infections that cause chronic hepatitis , liver cirrhosis , and hepatocellular carcinoma [7] . An estimated 170 million people are affected by the virus , making it a serious global health burden [8] . Recently , Gilead Sciences’ sofosbuvir/ledipasvir ( Harvoni ) and AbbVie's paritaprevir/ritonavir/ombitasvir plus dasabuvir ( Viekira Pak ) were approved as the new line of interferon-free treatment regimen . In addition , Miravirsen ( Santaris Pharma , Denmark ) , an antisense inhibitor of miR-122 , showed a decrease of HCV titers in patients chronically infected with HCV in phase II clinical trials [9] , demonstrating that miR-122 is a potential therapeutic host target to combat HCV . Here , we report an additional role for miR-122 in promoting HCV infection that is independent of its well-characterized 5’ end stabilization function . Like many RNA viruses , HCV exploits membranes and the trafficking machinery of the host for viral replication [10 , 11] . For example , accumulating evidence suggests that HCV can exit infected cells via the multivesicular transport system [12–15] . While these studies employed fractionation and ultrastructural approaches , evidence for the cellular origin or the mechanism of vesicle generation remains lacking . Recently , it has been reported that Rab27a modulates exosome vesicle secretion by docking multivesicular bodies to the plasma membrane [16] . Curiously , several studies have shown that Rab27a , a small GTPase , is also involved in replication of viral genomes in cells infected with human immunodeficiency virus , herpes simplex virus , hepatitis E virus and HCV [15 , 17–19] . However , the mechanism by which Rab27a modulates viral genome replication remains unclear . In this study , we found that Rab27a affects HCV RNA and virion abundance by a pathway that is independent of exosome secretions . Specifically , Rab27a located to membranes that are enriched in viral replication complexes and to lipid droplets , which are sites thought to initiate packaging of the viral RNA genome . Furthermore , intracellular abundance of Rab27a affected miR-122 abundance . Curiously , Rab27a’s modulation of miR-122 was independent of miR-122’s stabilizing role of the viral RNA . Therefore , Rab27a likely downregulates , via miR-122 , a cellular inhibitor of HCV gene expression .
To determine whether HCV RNA and protein abundances are regulated by exosomal vesicles , we first inhibited exosomal trafficking in human liver carcinoma Huh7 cells by depletion of Rab27a [16] . Northern analyses revealed that the liver Rab27a gene is transcribed into three RNA transcripts of 1 . 2 kb , 2 . 6 kb and 3 . 5 kb in size ( Fig 1A ) . This result is consistent with Rab27a RNA species that are expressed in human fibrosarcoma cells [20] . All three Rab27a transcripts were decreased by 90% in both uninfected and JFH1 HCV-infected cells that were treated with siRNAs directed against the common Rab27a open reading frame ( Fig 1A and 1B ) . As expected , Western blot analysis showed that the abundance of Rab27a protein was also decreased in Rab27a siRNA-treated cells ( Fig 1D ) . To determine if Rab27a depletion affected extracellular exosome yield , the abundance of CD81 , a marker for exosomes derived from the multivesicular body pathway , was examined in cell lysates and in extracellular , partially purified exosome preparations . S1 Fig shows that Rab27a depletion diminished the extracellular amount of CD81-containing exosomes in uninfected ( S1A Fig ) and in HCV-infected cells by approximately 40% ( S1B Fig ) . To examine the effects of Rab27a on HCV gene expression , viral RNA and protein abundances were measured in Rab27a-depleted cells . Results showed that Rab27a depletion caused a 60% decrease in HCV RNA abundance ( Fig 1A , lane 4 , and Fig 1C ) , but had no effect on actin mRNAs . Rab27a depletion also led to a decrease in HCV core protein abundance ( Fig 1D ) . These data are consistent with a previous report on the effect of Rab27a depletion on HCV RNA abundance [15] . Similar effects of decreased viral RNA ( S2A Fig ) and protein ( S2B Fig ) protein abundances during Rab27a depletion were observed when cells were infected at a 1000-fold higher multiplicity of infection with HCV . To control for siRNA off-targeting effects , additional Rab27a siRNAs ( siRNA-3 and siRNA-4 ) , which target different regions of all Rab27a mRNA species , were tested . These siRNAs also showed decreased Rab27a and HCV RNA ( S3A Fig ) and protein abundances ( S3B Fig ) . Importantly , Rab27a depletion in Huh7 cells did not have a significant effect on cell viability ( S4A Fig ) or caused apoptosis ( S4B Fig ) . Therefore , depletion of Rab27a causes selective inhibition of HCV gene expression without any significant effects on cellular viability . It has been reported that HCV can be transmitted from cell to cell via exosomes [12 , 14 , 21–23] . Rab27a plays a role in exosome secretion . Thus , we would expect an increase in cell-associated virus titer in Rab27a-siRNA treated cells compared to control-siRNA treated cells . Depletion of Rab27a decreased extracellular virus titer by about 80% ( Fig 2A ) , but , surprisingly , cell-associated virus titer also decreased by about 60% ( Fig 2B ) . However , the ratio of cell-associated to total infectious virus particles in Rab27a-depleted cells was similar to that of control-siRNA treated cells ( Fig 2C ) . Consistently , extracellular HCV RNA abundance was decreased to nearly 80% in infected cells that were treated with Rab27a siRNAs , compared to Ctrl siRNAs-treated cells ( 4 . 9 x 106 copies/ml ) ( Fig 2D ) . The decrease of extracellular HCV RNA abundance did not cause an accumulation of intracellular HCV ( Fig 1A , lane 4 , and Fig 1C ) . Thus , these data suggest that the diminished yield of cell-associated infectious virus particles during Rab27a depletion is not due to impaired exosome secretion , arguing that Rab27a modulates HCV gene expression by a mechanism that is different from its role in exosome secretion . To determine whether Rab27a modulates viral RNA abundance at the RNA replication or translation step , and to bypass any effects on virion entry , we monitored the expression of subgenomic JFH1-Rluc ( sgJFH1-Rluc ) replicons [24 , 25] ( Fig 3A ) . These replicons are either competent for both translation and RNA replication , or contained a GND mutation in the catalytic domain of the viral RNA-dependent RNA polymerase ( NS5B ) that prevents genome replication ( Fig 3A ) . Briefly , Huh7 cells were transfected with Rab27a siRNAs , and subsequently transfected with replication-competent sgJFH1-Rluc RNAs ( Fig 3B ) or replication-defective sgJFH1-Rluc-GND RNAs ( Fig 3C ) . Luciferase activity was measured at different times after HCV RNA transfection . Two peaks of luciferase activity were noted in the sgJFH1-Rluc RNA-transfected cells treated with control siRNAs ( Fig 3B ) . The first peak at 4 hours post-transfection represents the initial translation of the input RNA , which is absent in cyclocheximide-treated cells ( Fig 3B and 3C ) . The second luciferase peak represents the translation of replicating RNAs , because it is absent in sgJFH1-Rluc-transfected cells that were treated with the NS5B inhibitor MK-0608 ( Fig 3B ) and in sgJFH1-Rluc-GND-transfected cells ( Fig 3C ) . Depletion of Rab27a did not diminish translation of the input RNA ( Fig 3B and 3C ) . However , translation of replicating RNAs was significantly decreased in Rab27a-depleted cells compared to control siRNA-treated cells . Importantly , the EMCV IRES activity was not affected by Rab27a depletion ( S5 Fig ) , eliminating the possibility that these results were due to altered abundances of viral proteins . These findings argue that Rab27a plays a role in the viral life cycle by modulating HCV RNA replication . It is known that cells expressing HCV replicons or cells that are infected with HCV display membrane rearrangements and formation of virus-induced membranous webs [11 , 26–29] . The HCV-induced membranous webs , which are thought to be the sites of viral replication , are mainly derived from the endoplasmic reticulum ( ER ) [29] . To examine whether Rab27a is located to membranes during HCV RNA replication , membrane-enriched fractions from uninfected and HCV-infected cells were isolated , using discontinuous sucrose gradients . Western blot analyses showed that the membrane fractions contained the ER membrane marker protein calnexin ( Fig 4A and 4B , lanes 3 and 4 ) . In addition , HCV proteins NS5A , NS3 and capsid protein core also located to these fractions ( Fig 4A and 4B ) . Interestingly , Rab27a was also found to localize in the membrane-enriched fraction . Rab27a depletion caused a decrease of HCV NS3 , NS5A and core protein abundance in the enriched-membrane fraction ( Fig 4B , lane 4 ) , but not calnexin or GAPDH . These results indicate that Rab27a is associated with membrane-enriched fractions in infected cells , and that Rab27a depletion selectively diminished the abundance of several viral non-structural proteins in the replication complex-containing membranes . The above genetic and biochemical findings argue that Rab27a regulates HCV RNA replication via its association with virus-induced membranes . To further substantiate this hypothesis , the subcellular location of Rab27a was investigated by confocal immunofluorescence microscopy . Astonishingly , Rab27a exhibited a doughnut-like structural localization around lipid droplets ( LDs ) ( Fig 5 and S6 Fig ) in uninfected ( Fig 5A ) and in infected liver cells ( Fig 5B ) . These findings suggest that Rab27a may have a hitherto unknown role in the metabolism of LDs in liver cells . The LD-Rab27a doughnut-like structures colocalized with viral core protein in infected cells ( Fig 5B ) . In addition , a small fraction of NS3 displayed a punctate distribution in the LD-Rab27a structures , indicating that Rab27a localizes to adjacent to sites of viral replication ( S6B Fig ) . The impaired HCV gene and protein expression may be due to a lack of stabilization of HCV RNA . To examine whether Rab27a affects HCV RNA stability , Huh7 cells were transfected with control- or Rab27a-siRNAs , followed by addition of the NS5B inhibitor MK-0608 to block new synthesis of HCV RNA . The rate of HCV RNA decay was determined by Northern blot analysis at different times after addition of MK-0608 ( Fig 6A ) . Viral RNAs from control- and Rab27a-depleted samples displayed similar decay rates , with approximate half-lives of 4 . 8 hours ( Fig 6B ) . These results indicate that Rab27a depletion affects the rate of HCV RNA replication without changing HCV RNA stability . It is known that miR-122 modulates HCV RNA expression [3 , 30] . Therefore , it is possible that the observed effects of Rab27a depletion on the rates of HCV RNA replication could be due to altered abundance of miR-122 . Thus , intracellular miR-122 abundance was monitored in Rab27a-depleted cells by Northern blot analysis . Results showed that miR-122 abundance was decreased by more than 30% in both uninfected- and HCV-infected Rab27a-depleted cells ( Fig 7A and 7B ) . This was surprising because miR-122 has been reported to be quite stable in liver cells [31] . A luciferase reporter-based assay also showed diminished miR-122 function in Rab27a-depleted cells ( S7 Fig ) . While the abundances of five other endogenous miRNAs ( miR-16 , miR-21 , miR-22 , miR-26 and miR-130a ) were not changed in uninfected , Rab27a-depleted cells ( Fig 7A ) , the abundances of miR-16 , miR-22 and miR-130a showed a modest decrease in Rab27a-depleted cells during HCV infection; but not to the same extent as miR-122 ( Fig 7B ) . To test whether the modulation of HCV RNA replication by Rab27a was caused by the altered abundance of miR-122 or any other microRNA , we investigated whether miR-122 overexpression prevented the Rab27a-dependent inhibition of HCV RNA replication ( Fig 8A ) . Fig 8B shows that overexpression of miR-122 mimetics could rescue HCV RNA abundance in Rab27a-depleted cells , while the overexpression of miR-22 had no effects . A similar result was observed during overexpression of miR-21 as a control . These findings suggest that the decrease of HCV RNA abundance in Rab27a-depleted cells is due to the reduction in miR-122 abundance and is not due to the reduction of other microRNAs , such as miR-22 ( Fig 8B ) . We next examined whether Rab27a modulates the transcription of miR-122 . Primary miR-122 ( pri-miR-122 ) transcript abundance was examined in Rab27a-depleted uninfected or HCV-infected cells . S8A and S8B Fig shows that the abundance of pri-miR-122 is not affected by the depletion of Rab27a in uninfected and infected cells , suggesting that Rab27a modulates miR-122 abundance at a post-transcriptional step . Because precursor-miR-122 ( pre-miR-122 ) can not be detected in cultured Huh7 cells , we determined the effect of Rab27a on the stability of a pre-miR-122 species that is resistant to the cleavage by Dicer [32] . Thus , the intracellular decay of a dicer-resistant pre-p3 ( dNx12 ) that is functional in regulating mRNAs with miR-122 target sites [32] was examined ( S9A Fig ) . Control- or Rab27a-siRNA treated cells were transfected with 5’-32P-labelled pre-p3 ( dNx12 ) mimetics and the abundance of the labeled pre-miRNAs was determined at one day after transfection . The three independent experiments in S9B Fig show that the abundance of 5’-32P-labelled pre-p3 ( dNx12 ) significantly decreased by the depletion of Rab27a ( S9C Fig ) , arguing that Rab27a likely diminished miR-122 abundance by decreasing pre-miR-122 abundance . It is known that two miR-122 molecules protect the 5’-terminal sequence of the HCV RNA genome from exonucleolytic degradation [5 , 6] . Thus , it was possible that the reduced level of intracellular miR-122 , after Rab27a depletion , caused the decrease in HCV RNA abundance by leaving the viral RNA unprotected . To test this possibility , a mutant HCV RNA genome ( HCV-G27G42 ) that contained a mutation at each of the two miR-122 binding sites at the 5’ UTR was generated ( Fig 9A , nucleotides highlighted in red ) . When transfected into cells , HCV-G27G42RNA cannot replicate because it cannot bind endogenous miR-122 ( Fig 9A , ( I ) , left upper panel ) [3 , 4 , 30] . However , introduction of p3-loop miR-122 molecules that harbor a compensatory mutation at position 3 ( red ) , and additional mutations at positions 9–13 ( orange ) and 18 ( orange ) ( Fig 9A , ( I ) , lower panel ) can enhance HCV-G27G42 RNA abundance ( Fig 9B , lanes 1 and 2 ) . The nucleotide changes 9–13 ( orange ) and 18 ( orange ) in p3-loop miR-122 allow us to distinguish p3-loop miR-122 from endogenous wildtype miR-122 in Northern blots . As a negative control , miR-122 molecules with mutations in their entire seed sequences ( p2-8; nucleotides highlighted in blue ) ( Fig 9A , ( II ) , lower panel ) did not enhance HCV-G27G42 RNA abundance ( Fig 9B , lane 4 ) . This finding shows that the HCV-G27G42 RNA genome abundance was enhanced by p3-loop miR-122 , and not by endogenous miR-122 or p2-8 miR-122 . Expression of p3-loop miR-122 mimetics allowed a 50% of HCV RNA accumulation in Rab27a-depleted cells ( Fig 9B , lane 3 ) compared to cells that were not depleted of Rab27a ( Fig 9B , lanes 1 and 2 ) . Quantitation of the abundances of the endogenous and p3-loop miR-122 molecules revealed that endogenous miR-122 abundance was diminished by 30% in Rab27a-depleted cell ( Fig 9C ) , a finding that is consistent with the result in Fig 7A and 7B . In contrast , the abundance of p3-loop miR-122 was not affected by Rab27a depletion ( Fig 9C ) . Therefore , the 50% decrease in HCV-G27G42 RNA abundance in Rab27a-depleted cells in the presence of p3-loop miR-122 mimetics ( Fig 9B , lane 3 ) , is independent of the interaction of p3-loop miR-122 with the 5’ end of HCV RNA . This findings argue that endogenous miR-122 , but not p3-loop miR-122 , downregulates the expression of an inhibitor of HCV RNA gene expression .
CD81-containing exosomes are multivesicular body-derived microvesicles found in eukaryotic cells and are involved in cell-to-cell communication . It has been shown that both mRNAs and miRNAs can be transferred into neighboring cells by this pathway [33] , and that HCV RNA can also be secreted from infected cells by extracellular vesicles [12 , 14 , 21–23] . However , extracellular vesicles , including exosomes , can be derived from several distinct pathways . To test whether HCV RNA and miR-122 are secreted by bona-fide exosomes , Rab27a that modulates the docking of multivesicular bodies to the plasma membrane [16] was depleted by siRNAs . Indeed , depletion of Rab27a led to a decrease of CD81- and CD63-positive exosome secretion in Huh7 cells ( S1 Fig ) . Previous studies have argued that viral RNA can be transferred by “exosomes” [12 , 14 , 21–23] , which were isolated from supernatants of cultured cells by subsequent centrifugation steps and CD81 affinity chromatography . In contrast , we show here that depletion of exosomes by genetic downregulation of the exosome docking protein Rab27a lowered both the intracellular and extracellular abundance of HCV RNA and virions ( Fig 2 ) , arguing that microvesicles other than exosomes are the major vehicles for the transport of viral RNA and virions . The effect of Rab27a siRNA-3 , which targets the 3’ noncoding region of all Rab27a mRNAs , on HCV RNA abundance could not be restored by overexpressing a knockdown-resistant Rab27a variant . We also found that overexpression of Rab27a did not increase HCV RNA and extracellular exosome abundance . Thus , Rab27a may affect HCV RNA abundance and exosome secretion as part of a protein complex . Alternatively , siRNA-3 caused off-target effects that were unrelated to Rab27a . To examine the latter possibility , additional siRNAs targeting different regions of Rab27a mRNAs were tested . All siRNAs showed a decrease in HCV RNA abundance , supporting the specificity of Rab27a’s effect on HCV RNA abundance ( S3 Fig ) . Importantly , all siRNAs directed against Rab27a did not affect cell viability . Studies with HCV replicons provided genetic evidence that Rab27a modulates the rate of viral replication ( Fig 3 ) . To further substantiate this finding with a biochemical approach , we examined the protein composition of membranes , which are sites for viral RNA replication . A substantial amount of Rab27a located to membrane-enriched fractions , both in uninfected and infected cells ( Fig 4 ) . In addition , confocal microscopy studies revealed that Rab27a localizes to LDs in uninfected and infected cells ( Fig 5 ) . Curiously , Rab27a coats LDs , visualizing the Rab27a-LDs complex as a doughnut-shaped structure . LD-associated Rab27a colocalized with viral core protein and with a small fraction of NS3 . It has been proposed that HCV core recruits ER-derived membrane webs that are close to LDs to create a local membrane environment for viral replication and assembly [34 , 35] . While the exact mechanism by which Rab27a modulates HCV RNA abundance is not clear at present , our findings strongly argue that Rab27a regulates HCV RNA abundance at LDs . It is known that the presence of miR-122 is essential to maintain HCV RNA abundance . Profiling of several microRNAs in Rab27a-depleted cells showed that the abundance of miR-122 was decreased both in uninfected and infected cells . The loss of HCV RNA abundance during Rab27a depletion could be rescued by overexpression of miR-122 mimetics , which is consistent with the hypothesis that Rab27a-mediated depletion of miR-122 caused loss of HCV RNA abundance ( Fig 8 ) . No significant decrease in the amount of primary miR-122 was observed in Rab27a-depleted cells , indicating that the effect of Rab27a depletion on miR-122 most likely occurred at a post-transcriptional step in the cytosol . Indeed , Rab27a depletion caused a decrease of ectopically expressed precursor miR122 ( S9 Fig ) . It has been reported that both pre-microRNAs and mature microRNAs can be released from cells via exosomes that contain the GW182 component of the RNA-induced silencing complex ( RISC ) [36 , 37] . This observation raises the possibility that depletion of Rab27a enhances the intracellular abundance of GW182-containing vesicles that affect the stability of pre-miR122 or miRNA-122 molecules . However , depletion of Rab27a effector Slp4 did not affect miR-122 and HCV RNA abundances . Because depletion of Slp4 inhibits exosome trafficking [16] , loss of HCV RNA and miR-122 was not due to the accumulation of intracellular exosomes . We hypothesize that pre-miR-122 is being destabilized in the absence of Rab27a by an as-of-yet unknown mechanism . We also noted a selective decrease of several microRNAs in infected cells . One explanation is that HCV infection causes a dispersion of Processing bodies , where microRNAs , microRNA-targeted mRNAs and Argonaute proteins are located [38 , 39] . This dispersion may affect turnover of specific microRNAs in infected cells . Alternatively , HCV is known to sequester components of RISC , such as Ago2 and GW182 , at the HCV 5’ end for maintaining viral genome stability [40] . As a consequence , RISC-free microRNAs may be more easily degraded [22 , 41 , 42] . It is important to note that both miR-122 and miR-22 are depleted in HCV-infected cells . However , only the depletion of miR-122 affects HCV RNA abundance ( Fig 8 ) , arguing that loss of HCV RNA abundance was not caused by an overall loss of microRNAs in infected cells . We examined whether loss of miR-122 led to the accumulation of HCV-G27G42 RNA molecules that were vulnerable to exonuclease cleavage . Thus , we examined the abundance of HCV-G27G42 RNA that could be protected by ectopically expressed mutant miR-122 molecules in Rab27a-depleted cells . The abundance of HCV-G27G42 RNA that could interact with mutant miR-122 , but not with endogenous miR-122 , also decreased in Rab27a-depleted cells ( Fig 9 ) . Because mutant miR-122 molecules very likely do not recognize mRNA targets that are modulated by wildtype , endogenous miR-122 , effects of endogenous miR-122 on HCV RNA abundance are by a mechanism that is different from its protecting the 5’ end of the viral RNA . We also examined whether a miR-122 antagonist , instead of Rab27a depletion , caused a decrease in miR-122 to affect HCV replication that is independent of endogenous miR-122 . We noted to our surprise that exogenously expressed mutant miR-122 mimetics cannot be functionally sequestered by the employed antagomirs . Thus , it is possible that an antagomir-inaccessible pool of mutant miR-122 accumulates within the transfected cell . Finally , depletion of Rab27a has no effect on exoribonucleases Xrn1 and Xrn2 abundance . Thus , it is very likely that miR-122 downregulates an inhibitor of HCV gene expression . Such an inhibitor is not involved in the biosynthesis of cholesterol , because cholesterol abundance is not affected in Rab27a-depleted uninfected or infected cells .
Human hepatoma Huh7 cells were kindly provided by Francis V . Chisari ( The Scripps Research Institute , San Diego ) . Huh7 cells were cultured in DMEM supplemented with 10% fetal bovine serum , 1x non-essential amino acids and 2 mM L-glutamine ( Gibco ) . All Small interfering RNA ( siRNA ) oligonucleotides and other RNA oligonucleotides were synthesized by Stanford PAN facility ( Stanford , CA ) . The siRNA sequences are as follow: siControl , 5’- GAUCAUACGUGCGAUCAGAdTdT-3’; siRab27a-1: 5’- GGAGAGGUUUCGUAGCUUAdTdT-3’; siRab27a-2: 5’- GCCUCUACGGAUCAGUUAAdTdT-3’ . The RNA oligonucleotide sequences are as follow: p3-loop miR-122: 5’- UGCAGUGUCUAUUUGGUCUUUGU-3’; p2-8 miR-122: 5’- UAAUCACAGACAAUGGUGUUUGU-3’ . For formation of RNA duplexes , 50 μM of sense and antisense strands were mixed in annealing buffer ( 150 mM HEPES ( pH 7 . 4 ) , 500 mM potassium acetate , and 10 mM magnesium acetate ) to a final concentration of 20 μM , denatured for 1 min at 95°C , and annealed for 1 h at 37°C . Huh7 cells ( 106 ) were seeded in 10 cm tissue culture dishes . Cells were infected with wild-type JFH1 at a MOI of 0 . 01 for 5 h , washed with PBS to remove unbound virus , trypsinized and replated in 15 cm tissue culture dishes . The supernatant was collected at 3 days post-infection and centrifuged at 1 , 000 rpm , 10 min at 4°C to remove cell debris . The infected cells were scraped and resuspended in medium and subjected to freezed-thraw cycles . Samples were centrifuged at 1 , 000 rpm , 10 min at 4°C to remove cell debris . For the virus stock , the supernatant was mixed with cell-associated virus . Virus was stored in aliquots at -80°C . Virus titter was determined by using fluorescent focus-forming assay . Huh7 cells ( 2 . 5 x 105 ) were seeded in 60 mm tissue culture dishes . Cells were transfected the following day with 50 nM of siRNA duplexes ( 25nM siRab27a-1 plus 25 nM siRab27a-2 ) using Dharmafect I reagent ( Dharmacon ) according to the manufacturer’s instruction . After 24 h post-transfection , the cells were infected with HCV JFH-1 virus at a MOI of 0 . 01 at 37°C . After 5 h incubation , cells were washed with PBS to remove unbound virus , trypsinized and replated in duplicate tissue culture dishes . Virus-infected cells were transfected again with 50 nM of siRNA duplexes at day 1 post-infection , and harvested at day 3 post-infection . The efficiency of siRNA depletion was evaluated by Northern and Western blot analysis . Huh7 cells were washed once with PBS and total RNA was extracted using TRIzol ( Invitrogen ) following the manufacturer’s protocol . Ten μg of total RNA in RNA loading buffer ( 32% formamide , 1x MOPS-EDTA-Sodium acetate ( MESA , Sigma ) and 4 . 4% formaldehyde ) was denatured at 65°C for 10 min and separated in a 1% agarose gel containing 1x MESA and 3 . 7% formaldehyde . The RNA was transferred and UV crosslinked to a Zeta-probe membrane ( Bio-Rad ) . The membrane was hybridized using the ExpressHyb hybridization buffer ( Clontech ) or ULTRAhyb ( Ambion ) and α-32P dATP-RadPrime DNA labelled probes ( Invitrogen ) complementary to HCV ( nucleotides 84–374 ) , Rab27a ( nucleotides 664–1145 ) , or actin ( nucleotides 685–1171 ) . Autoradiographs were quantified using ImageQuant ( GE Healthcare ) . Ten μg of total RNA was separated in 12% acrylamide/ 7 M urea gel . Small RNAs were transferred onto a Hybond-N+ membrane ( GE Healthcare ) , and detected by γ-32P-end labelled DNA probes complementary to miR-122 , miR-16 , miR-21 , miR-22 , miR-26 , miR-130a , mutant miR-122 or U6 snRNA . Oligonucleotide sequence of probes are: miR-122 probe , 5’-CAAACACCATTGTCACACTCCA-3’; miR-16-5p probe , 5’-CGCCAATATTTACGTGCTGCTA-3’; miR-21 probe , 5’-TCAACATCAGTCTGATAAGCTA-3’; miR-22-3p probe , 5’-ACAGTTCTTCAACTGGCAGCTT-3’; miR-26a-5p probe , 5’- AGCCTATCCTGGATTACTTGAA-3’; miR-130a-3p probe , 5’- ATGCCCTTTTAACATTGCACTG-3’; U6 probe , 5’-CACGAATTTGCGTGTCATCCTTGC-3’ . The membrane was hybridized using 7 . 5 x Denhardt’s solution , 5 x SSPE , 0 . 1% SDS , 0 . 05 mg/ml tRNA . Autoradiographs were quantified using ImageQuant ( GE Healthcare ) . Cells were washed with PBS once and lysed in RIPA buffer ( 50mM Tris ( pH8 . 0 ) , 150 mM NaCl , 0 . 5% sodium deoxycholate , 0 . 1% SDS , and 1% Triton X-100 ) containing Complete EDTA-free protease inhibitors ( Roche ) for 15 min on ice . The cell lysate was clarified by centrifugation at 14 , 000rpm for 5 min at 4°C . Forty μg of cell lysate was mixed with 2x SDS sample buffer ( 126 mM Tris HCl , 20% glycerol , 4% SDS and 10% β-mercaptoethanol , 0 . 005% bromophenol blue , pH 6 . 8 ) , denatured at 90°C for 5 min and separated in a 10% SDS-polyacrylamide gel . Protein was transferred to a PVDF membrane ( Millipore ) . The membrane was blocked with 5% non-fat milk in PBS-T and probed using primary antibody , followed by horse-radish peroxidase-conjugated secondary antibodies . The blot was developed using Pierce ECL Western Blot Substrate ( Thermo Scientific ) according to the manufacturer’s instructions , and exposed to Biomax Light Films . The following primary antibodies were used for western blot analysis: anti-Core ( C7-50 ) ( Abcam , ab2740 ) , anti-Rab27a ( Abnova ) , anti-GAPDH ( Calbiochem CB1001 ) . Infectious titers were determined by measuring fluorescent focus forming units ( FFU ) [43] . Rab27a depleted cells were infected with JFH-1 virus . For extracellular virus , supernatant of the infected cells was collected at day 3 post-infection . To harvest cell-associated virus , infected cells were washed with PBS three times , collected into a new tube , and resuspended in 500 μl DMEM . The cells were frozen and thawed three times . Both extracellular and cell-associated supernatants were sedimented at 14 , 000 rpm , 4°C for 5 min to remove cell debris . The viral titer was determined by FFU assay . Briefly , 3 . 2 x 104 cells were seeded in a 48-well plate and incubated overnight . A serial dilution of virus stock was added to cells and incubated for 5 h at 37°C . The diluted virus supernatant was removed from cells . Cells were washed with PBS and replaced with fresh medium . At day 3 post-infection , infected cells were washed once with PBS and fixed with cold methanol/acetone ( 1:1 ) . The level of HCV infection in the cells was analyzed by using a mouse monoclonal antibody direct against HCV core ( Abcam ) at 1:1000 dilution in 1% fish gelatin/PBS at 4°C overnight and an AlexFluor488- conjugated goat anti-mouse antibody ( Invitrogen ) at 1: 200 dilution at room temperature for 2 h . The fluorescent focus forming units were counted using a fluorescence microscope , and the viral titer was expressed as FFU per ml . Cell culture supernatants were collected from infected Huh7 cells . HCV RNAs from the supernatant were isolated using TRIzol LS reagent ( Inviterogen ) following the manufacturer’s protocol . HCV transcripts were quantified using SuperScript III Platinum SYBR Green One-Step qRT-PCR kit ( Invitrogen ) . The reactions were performed using the CFX connect Real-Time system ( BIO-RAD ) . HCV transcript levels were determined by comparison to standard curves derived from in vitro transcribed HCV RNA . The primer sequences for JFH1 were , Fwd , 5’-TCTGCGGAACCGGTGAGTA-3’; Rev , 5’-TCAGGCAGTACCACAAGGC-3’ . The plasmid H77ΔE1/p7 , containing a deletion of structural proteins E1-E2-p7 [44] was transcribed using the T7 MEGAscript kit ( Ambion ) , according to the manufacturer’s protocol . A mutant HCV RNA ( nucleotide 27 and 42 C to G change ) from H77ΔE1/p7-S1+2:p3 was transcribed as described [3 , 4 , 39] . Huh7 cells were transfected with Rab27a siRNAs ( 50 nM ) at day 1 and mutant miR-122 duplex ( 50 nM ) at day 2 . Subsequently , cells were electroporated with the mutant HCV RNA at day 3 . Briefly , Huh7 cells in 10cm dishes were trypsinized , washed with PBS once , and then washed with the Cytomix buffer , and suspended in the Cytomix buffer ( 120mM KCl , 0 . 15 M CaCl2 , 10mM K2HPO4 , 25 mM HEPES , 2 mM EDTA , 5 mM MgCl2 , pH7 . 6 ) , containing 10 μg HCV RNA . The cells were electroporated in 0 . 4 cm Biorad cuvette at 900V , 25 μF , and ∞ resistance , then incubated at room temperature for 10 min and seeded in a new 10cm dish . The cells were transfected again with Rab27a siRNAs and mutant miR-122 duplexes at 1 day after electroporation and harvested 3 days after electroporation . Subgenomic JFH1-Rluc and JFH1-Rluc-GND were kindly provided by Glenn Randall ( University of Chicago ) . The replicon RNA was generated using the T7 MEGAscript kit ( Ambion ) according to manufacturer’s protocols . Huh7 cells in 6 well plates were transfected with control or Rab27a siRNA using the Dharmafect I reagent ( GE Dharmarcon ) . After 1 day post-transfection , cells were transfected with 2 μg of replicon RNA in TransMessenger reagent ( Qiagen ) for 1 h , and replaced with complete medium according to manufacturer’s instructions . Cells were harvested at 1 , 2 , 4 , 8 , 12 , 24 , 36 and 48 hours . Luciferase activity from the sample was detected according to manufacturer’s instructions . Membrane-enriched fractions were isolated using a modified protocol adapted from Schlegel et al . [45] . Briefly , cells were washed with cold PBS twice , scraped in PBS , and pelleted . Cells were suspended in hypotonic buffer ( 10 mM Tris ( pH 8 . 0 ) , 10 mM NaCl , 1 mM MgCl2 , with complete protease inhibitor cocktail tablets ( Roche ) and 0 . 5 mM PMSF ) for 10 min on ice and then homogenized for 50 strokes using a Dounuce homogenizer . The cell homogenate was centrifuged at 1 , 000 x g for 10 min to remove nuclei and unbroken cells . The supernatant was collected and salt concentration was adjusted by adding NaCl to a final concentration of 300 mM . The cytoplasmic extract was then layered on a 10% and 60% sucrose in 300 mM NaCl , 15 mM Tris-HCl ( pH7 . 5 ) , 15 mM MgCl2 , and centrifuged at 26 , 000 rpm at 4°C in a SW41 rotor for 16 h . The viscous layer in the middle of the gradient was collected using a syringe . The sample was concentrated with a Nanosep 3K Omega centrifugal device ( Pall Life Sciences ) . The sample was resuspended in 4 x SDS sample buffer and separated in a 10% SDS-polyacrylamide gel . Uninfected and HCV-infected Huh7 cells were grown on 8-chambered coverglass slides ( LabTek II chamber slides , Thermo Scientific ) for 3 days . Cells were rinsed with PBS and fixed with 4% paraformaldehyde ( Electron Microscopy Sciences ) in PBS for 20 min at RT . Cells were then washed with PBS for 5 min twice , and permeabilized with 0 . 1% Triton X-100 in 1% fish gelatin ( Sigma ) in PBS ( 1% PBS-FG ) for 5 min . Blocking incubation was performed in 1% PBS-FG for 10 min , 3 times , at RT . Cells were incubated with primary antibodies in 1% PBS-FG at 4°C overnight , washed with 1% PBS-FG for 10 min twice , and incubated with secondary antibodies for 2h at RT . To visualize lipid droplets , cells were by stained with BODIPY 493/503 . After washing with 1% PBS-FG for 10 min twice , Hoechst 33258 dye ( Sigma ) in 1% PBS-FG was added and cells were incubated 5 min at RT . After two washes in 1% PBS-FG for 5 min each , the coverglass slides were embedded in Fluoromount-G ( SouthernBiotech ) . Samples were imaged at RT ( 22°C ) with a 20×/N . A . 0 . 60 or a 63×/N . A . 1 . 30 oil Plan-Apochromat objective on a Leica SPE laser scanning confocal microscope ( Leica-microsystems ) . Images were processed with ImageJ ( Ver . 1 . 48 , NIH ) using only linear adjustments of contrast and color . The following antibodies and reagents were used for immunofluorescence staining . Primary antibodies: mouse anti-Rab27a ( H00005873-M02 , Abnova ) , goat anti-HCV core ( 2861 , Virostat ) , goat anti-HCV NS3 ( 2871 , Virostat ) . Secondary antibodies: Alex Fluor 555 conjugated donkey anti-goat IgG ( H+L ) ( A-21432 ) and Alex Fluor 647 donkey anti-mouse IgG ( H+L ) ( A-31571 , Life technologies ) . Primary antibodies were used at 1:100 dilution and secondary antibodies were used at 1:200 dilution . Bodipy 493/503 was 1:100 dilution from 1 mg/ml stock ( D3922 , Invitrogen ) . Hoechst 33258 dye was 1:10 , 000 dilution from 2 mg/ml stock ( Sigma ) . Statistical analyses were performed with Prism 5 ( GraphPad ) . A two-tailed paired Student’s t-test was employed to assess significant differences between two groups . Error bars represent standard error of the mean . | Eukaryotic cells constantly expel a variety of small vesicles that are loaded with proteins , nucleic acids and other small compounds that were produced inside the cell . One particular kind of vesicle is called exosome . Exosomes are initially located in multivesicular compartments inside cells and are docked at the cell surface membrane by the small GTPase Rab27a . In the liver , high expression of Rab27a correlates with the development of hepatocellular carcinoma , suggesting a high trafficking capacity for exosomes . Also , it has been shown that hepatitis C virus ( HCV ) can spread from cell to cell via exosomes . We discovered that Rab27a abundance affects HCV virion abundance that independent from its role in exosome secretion . The presence of Rab27a in membrane-enriched replication complexes and nearby lipid droplets points to functions of Rab27a in the viral life cycle . Depletion of Rab27a resulted in a lower abundance of the liver-specific microRNA miR-122 . It is known that two molecules of miR-122 form an oligomeric complex with the 5’ end of the viral RNA leading to protection of the viral RNA against cellular nucleases . However , we show that the Rab27a-mediated loss of miR-122 was independent of its role in protecting the viral RNA , very likely by the downregulation of a cellular inhibitor of HCV gene expression . These findings argue for novel , hitherto undetected roles for miR-122 in the viral life cycle . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Supporting Role for GTPase Rab27a in Hepatitis C Virus RNA Replication through a Novel miR-122-Mediated Effect |
In order to control malaria , it is important to understand the genetic structure of the parasites in each endemic area . Plasmodium vivax is widely distributed in the tropical to temperate regions of Asia and South America , but effective strategies for its elimination have yet to be designed . In South Korea , for example , indigenous vivax malaria was eliminated by the late 1970s , but re-emerged from 1993 . We estimated the population structure and temporal dynamics of transmission of P . vivax in South Korea using microsatellite DNA markers . We analyzed 255 South Korean P . vivax isolates collected from 1994 to 2008 , based on 10 highly polymorphic microsatellite DNA loci of the P . vivax genome . Allelic data were obtained for the 87 isolates and their microsatellite haplotypes were determined based on a combination of allelic data of the loci . In total , 40 haplotypes were observed . There were two predominant haplotypes: H16 and H25 . H16 was observed in 9 isolates ( 10% ) from 1996 to 2005 , and H25 in 27 ( 31% ) from 1995 to 2003 . These results suggested that the recombination rate of P . vivax in South Korea , a temperate country , was lower than in tropical areas where identical haplotypes were rarely seen in the following year . Next , we estimated the relationships among the 40 haplotypes by eBURST analysis . Two major groups were found: one composed of 36 isolates ( 41% ) including H25; the other of 20 isolates ( 23% ) including H16 . Despite the low recombination rate , other new haplotypes that are genetically distinct from the 2 groups have also been observed since 1997 ( H27 ) . These results suggested a continual introduction of P . vivax from other population sources , probably North Korea . Molecular epidemiology using microsatellite DNA of the P . vivax population is effective for assessing the population structure and transmission dynamics of the parasites - information that can assist in the elimination of vivax malaria in endemic areas .
Plasmodium vivax , the second most prevalent species of the human malaria parasite , is widely distributed around the world , especially in Asia and South America; it ranges from tropical to temperate areas [1] , [2] . In these countries , the proportion of P . falciparum cases is gradually decreasing due to the impact of global malaria control programs such as “The Roll Back Malaria Partnership” and “The Global Fund to Fight AIDS , Tuberculosis and Malaria” as well as local control programs . In contrast , the proportion of P . vivax cases is gradually increasing [1] , and therefore deserves more attention than it has previously received [3] . Understanding the genetic characteristics of the malaria parasite population is important for monitoring the transmission pattern and evaluating the effectiveness of malaria control in endemic areas [4]–[7] . Recently , the population structure and transmission dynamics of P . vivax have been reported in some tropical and subtropical areas where the parasites are prevalent throughout the year or seasonally prevalent but not discontinuous during the year [8]–[13] . However , little is known about these characteristics in temperate areas where vivax malaria is only seasonally prevalent and discontinuous during the year . In the Republic of Korea ( South Korea ) , which is in the temperate zone of the continent of Asia , indigenous vivax malaria had been successfully eliminated by the late 1970s thanks to an effective program conducted by the National Malaria Eradication Service of the South Korean government with the support of the WHO [14]–[16] , but has re-emerged since 1993 [17] . At the beginning of the re-emergence , the patients were only South Korean soldiers , veterans , and soldiers from the US military who were serving in the border area between North and South Korea in the western Demilitarized Zone ( DMZ ) [18]–[20] . Gradually , however , the number of infected civilians who lived in or near the area increased [18] , suggesting local transmission of P . vivax between humans and Anopheles mosquitoes in the country . The number of vivax malaria cases increased steadily until 2000 ( 4 , 183 cases ) , then began to decrease gradually until 2004 ( 864 cases ) ( Fig . 1 ) [1] , [16] . In spite of continuous malaria control measures implemented by the South Korean government , the numbers of reported cases fluctuated between 1 , 000 and 2 , 000 cases per year from 2005 to 2009 [1] . The WHO reports that vivax malaria was more prevalent in the Democratic People's Republic of Korea ( North Korea ) , where there were 296 , 540 cases in 2001 and 14 , 845 cases in 2009 [1] , [19] . We previously conducted genetic epidemiological surveys of the P . vivax population in South Korea using DNA sequences of some antigenic molecules of the parasite ( circumsporozoite protein , Duffy binding protein , apical membrane antigen 1 , merozoite surface protein-1 ) and found that there were 2 genotypes in the country's parasite population [21]–[25] . The advantage of using antigenic molecules of the parasites for genetic epidemiology is that they could be vaccine candidates; however such antigenic molecules are under strong selective pressure from the host immune system , so the variation in the molecules might be biased due to this [26] . In previous studies , the isolates that were used were collected from vivax malaria patients in a single year so temporal changes in the parasite population could not be examined . In the present study , we examined the population structure and the transmission dynamics of P . vivax in South Korea temporally using 10 highly polymorphic neutral DNA markers of the parasite collected from 1994 to 2008 and compared these characteristics with those reported in tropical and subtropical areas . Based on these data , we provide a possible explanation as to why it has not been possible to eliminate vivax malaria in South Korea in spite of a continuous governmental effort .
A total of 255 P . vivax samples isolated from South Korean soldiers or veterans who had served in the DMZ from 1994 to 2008 were used in this study . These patients were also diagnosed by microscopic examination of peripheral blood smears when they contracted malaria . The patient blood samples were collected and preserved at −30°C until use . This study was performed according to the Ethical Guidelines for Clinical Research issued by the Ministry of Health , Labour and Welfare of Japan on July 31 , 2008 , and the Ethical Guidelines for Epidemiological Research issued by the Ministries of Health , Labour and Welfare , and of Education , Science , Culture , and Sports of Japan on December 1 , 2008 . Because of the long-term prior collection of widely distributed samples , written or oral informed consent from the patients for the specific purpose of this study could not be obtained at each sample collection . However , no author of the study was involved in gathering patient samples and the individual information of the donors was disconnected from the authors . Thus , all the samples were anonymized , and indeed it is most unlikely that the results obtained from the analysis of the isolated parasites would result in a breach of donor privacy . Parasite DNA was extracted from frozen whole blood samples by phenol-chloroform extraction after proteinase K digestion [27] or by QIAamp DNA Mini Kit ( Qiagen , Valencia , CA , USA ) . Ten microsatellite DNA loci were amplified by PCR . The loci were as follows: MS1 ( chromosome 3 ) , MS4 ( chromosome 6 ) , MS5 ( chromosome 6 ) , MS6 ( chromosome 11 ) , MS7 ( chromosome 12 ) , MS8 ( chromosome 12 ) , MS9 ( chromosome 8 ) , MS12 ( chromosome 5 ) , MS15 ( chromosome 5 ) and MS20 ( chromosome 10 ) . The PCR primer sets and amplification conditions were consistent with the protocol of Karunaweera et al . [28] . Sizes of fluorescently-labeled PCR products were measured on an Applied Biosystems Prism Genetic Analyzer 3130xl using GeneMapper ( R ) version 4 . 1 with a 500 ROX size standard ( Applied Biosystems , CA , USA ) . Amplified different-sized PCR products using the same primer sets were considered to be individual alleles within a locus , as size variation among isolates is consistent with the repeat number in a microsatellite locus [5] . The electropherogram shows peak profiles for the microsatellite loci , based on the fluorescence intensity of the labeled PCR products in this analysis . Multiple alleles per locus were scored if minor peaks were taller than at least one-third the height of the predominant allele for each locus . Multiple-genotype infections ( MGIs ) were defined as those in which at least one of the 10 loci contained more than one allele [5] . Population genetic analyses were performed based on allele frequencies of the 10 microsatellite loci of the population . The level of genetic diversity of the P . vivax population in South Korea was assessed by allele number per locus ( A ) and expected heterozygosity ( HE ) . HE values for each locus were calculated using HE = [n/ ( n−1 ) ] [1−Σpi2] , where n corresponds to the number of isolates examined and pi is the frequency of the ith allele . The statistical differences among those values were evaluated by Welch's t-test . Multilocus linkage disequilibrium ( LD ) was assessed using the standardized index of association ( IAS ) [29] , [30] . This analysis was performed using the LIAN 3 . 5 Web interface [31] . IAS was calculated using the formula IAS = ( VD/Ve−1 ) / ( l−1 ) with permutation testing of the null hypothesis of complete linkage equilibrium ( IAS = 0 ) , where VD is the observed mismatch variance , Ve is the expected mismatch variance , and l is the number of examined loci . Significances of the observed IAS values were calculated by Monte-Carlo simulation , using 10 , 000 random permutations of the data . This statistic is a variation of the method proposed by Maynard-Smith et al [29] . The results were standardized by the number of loci , to enable a comparison of different data sets [30] . This test was applied to the data sets from each population in two ways . First , the mixed-clone infections were excluded so that only the single-clone infections were analyzed , giving absolute confidence in the haplotype profile . Second , any multilocus genotype found in more than one isolate was only counted once in the analysis , i . e . unique haplotypes only , reducing the sample size slightly and thereby removing the possible effect of recent epidemic expansion of particular clones [5] . Microsatellite haplotypes of the isolates were determined based on a combination of the allelic data of the 10 loci . The relationships among the haplotypes were estimated by eBURST analysis [32] .
In the 10 loci , the number of alleles ( A ) for each locus was 2 to 7 ( average: 4 . 3 ) . The expected heterozygosity ( HE ) for each of these loci was 0 . 05 to 0 . 66 ( average: 0 . 43 ) ( Table 2 ) . Next , the P . vivax population was divided into 2 groups: one comprised of the 47 isolates collected from 1994 to 2000 when the numbers of vivax malaria cases increased; the other comprised of the 40 isolates collected from 2001 to 2008 , when the numbers of cases decreased until 2004 and then increased slightly . The level of genetic diversity was reassessed for each group . For the first group , the averages ± SE of A and HE were 2 . 70±0 . 26 and 0 . 36±0 . 06 , respectively . For the second group , the averages ± SE of A and HE were 3 . 80±0 . 57 and 0 . 50±0 . 10 , respectively ( Fig . 1 ) . The levels of genetic diversity were relatively higher in the second group , with P values at 0 . 11 and 0 . 24 for average A and average HE , respectively . Furthermore , we also divided the population into 3 groups , each covering 5-year periods: 1994 to 1998 ( 33 isolates ) , 1999 to 2003 ( 36 isolates ) and 2004 to 2008 ( 18 isolates ) . The level of genetic diversity was reassessed for each group ( Fig . 3 ) . For the first group , the averages ± SE of A and HE were 2 . 50±0 . 27 and 0 . 31±0 . 05 , respectively . For the second group , the averages ± SE of A and HE were 3 . 00±0 . 42 and 0 . 42±0 . 09 , respectively . For the third group , the averages ± SE of A and HE were 3 . 80±0 . 57 and 0 . 56±0 . 10 , respectively . The levels of genetic diversity gradually increased with P values at 0 . 06 and 0 . 05 if we compared the difference of the average A between the first group ( 1994–1998 ) and the third group ( 2003–2008 ) and the difference of the average HE between the first and the third group , respectively . Likewise , the analysis of genetic diversity , IAS values were also calculated for the two populations: one comprised the isolates collected from 1994 to 2000 and the other comprised the isolates collected from 2001 to 2008 , with permutation testing of the null hypothesis of IAS = 0 ( equilibrium of multilocus frequencies ) ( Table 3 ) . When the single-clone haplotype was used in the analysis , the IAS values of the former ( 1994–2000 ) and the latter ( 2001–2008 ) were 0 . 529 and 0 . 218 , respectively , whereas when the unique haplotypes were used in the analysis , those of the former and the latter were 0 . 346 and 0 . 173 , respectively . Significant linkage disequilibrium was observed in both populations ( P<0 . 001 ) . Similar to the analyses of genetic diversity , we also divided the population into 3 groups covering 5-year periods: 1994 to 1998 ( 33 isolates ) , 1999 to 2003 ( 36 isolates ) and 2004 to 2008 ( 18 isolates ) . The IAS values were also calculated for each group . When the single-clone haplotype was used in the analysis , the IAS values of the first ( 1994–1998 ) , the second ( 1999–2003 ) and the third ( 2004–2008 ) groups were 0 . 584 , 0 . 315 and 0 . 140 , respectively , whereas when the unique haplotypes were used in the analysis , those of the first , the second and the third were 0 . 408 , 0 . 231 and 0 . 153 , respectively ( Table 4 ) . Significant linkage disequilibrium was observed in both populations ( P<0 . 001 ) . Microsatellite haplotypes of the 87 isolates were determined based on a combination of the allelic data of the 10 microsatellite loci; 40 haplotypes ( H1–H40 ) were observed ( Table 5 ) . There were 2 major haplotypes ( H16 and H25 ) : H16 was observed in 9 isolates ( 10% ) out of the 87 isolates in samples collected from 1996 to 2005; H25 was observed in 27 isolates ( 31% ) out of the 87 isolates in samples collected from 1995 to 2003 . H16 and H25 share only 3 alleles in the loci , MS8 , MS12 and MS20 , but those in 7 other loci were different from each other . The relationships among the 40 haplotypes were estimated by eBURST analysis [32] with the following criterion: when 2 isolates shared more than 7 identical loci out of the 10 loci , they were connected with a branch ( Fig . 4 ) . Again , two major groups were found: Group 1 was composed of 36 isolates ( 41% ) including the isolates with H25; Group 2 was composed of 20 isolates ( 23% ) including those with H16 . Some new or isolated haplotypes , namely H5 , H6 , H7 , H8 , H11 , H12 , H19 , H31 , H32 , H33 , H35 , H36 , that were not included in the 2 major groups or connected to any other haplotypes , have also been observed since 1998 . H6 and H7 were not shown in Figure 2 because these haplotypes were quite different from the other haplotypes .
This is the first 15-year-long longitudinal study on P . vivax population genetics using highly polymorphic neutral markers . The present study demonstrated that the level of genetic diversity of the P . vivax population in South Korea was remarkably lower than the levels in tropical and subtropical areas reported by Karunaweera et al . [28] and Orjuela-Sánchez et al . [33] ( Table 2 ) . The 10 microsatellite loci used in the present study were a subset of the 14 loci used in the previous studies by other groups [28] , [33] . Imwong et al . also reported that the mean values of HE of P . vivax populations from Thailand ( n = 28 ) , India ( n = 27 ) and Colombia ( n = 27 ) were 0 . 77 , 0 . 76 and 0 . 64 , respectively [11] , using 11 other microsatellite loci in the genome . These values reported by Imwong et al . were also higher than those in South Korea . Sample size ( n ) and sampling conditions such as the size of sampling area and the length of sampling period may affect levels of genetic diversity of living organisms . In comparison to other studies the sample size of the present study ( n = 87 ) was relatively large and the sampling period ( 15 years: from 1994 to 2008 ) was relatively long [11] , [28] , [33] . Generally , one would expect to see an increase in the level of genetic diversity when these conditions ( a large number of samples and a long sampling period ) are present . However , the South Korean P . vivax population showed low levels of genetic diversity , suggesting that the effective size of the re-emerged P . vivax population in South Korea might be small . Microsatellite variation is strongly dependent on the length of repeat arrays [34] . Studies of numerous organisms have shown higher levels of variation in loci with long repeat arrays than those with short repeat arrays [35] . In the present study , however , even the locus with a long repeat array ( MS20 ) showed low levels of variation in the South Korean population . Also in the same population , MS8 and MS12 showed low levels of variation , although the loci were not short repeat arrays . The genetic diversities of the loci from Sri Lankan and Brazilian populations ( Table 2 ) were higher than those from the South Korean population . Therefore , the uniqueness of the diversity would not be solely dependent on the characteristics of the loci . In fact , mutations of microsatellite loci are generally considered to be neutral . However , if the loci are in a certain gene or close to a certain gene , the mutation may not be strictly neutral . Indeed , 6 of the 10 loci examined in this study ( MS4 , MS5 , MS8 , MS9 , MS15 , MS20 ) were in a gene coding a hypothetical protein or a known protein ( Table 2 ) . One of the 10 loci ( MS7 ) was between a gene coding a hypothetical protein and a gene coding a merozoite surface protein-7 which is expected to be under strong selective pressure . Therefore , the mutation of those 7 loci may not be strictly neutral . The allelic data suggested that the frequencies of strand-slippage events of the microsatellite loci during mitotic replication in the South Korean P . vivax population were very low because identical alleles in the known loci have been found for 10 years or longer in this population . Multiple genotype infection ( MGI ) is one of the important indexes of population genetics and epidemiology of malaria parasites because MGI is the first step in recombination of the parasite genome between different clones . In the case of P . falciparum , the rate of MGIs per population is basically associated with the endemicity [4] , [5] . That is , the MGI rate of P . falciparum population is higher in high transmission areas and lower in low transmission areas . However , this is not the case with P . vivax populations because high MGI rates were observed among the P . vivax populations in low transmission areas [11] , [12] . This feature could be attributed to relapse owing to hypnozoites in the liver of a vivax malaria patient . Although MGI is an important index , the methods or criteria of determining MGI is problematic . When any locus of the 10 loci showed more than 1 allele , we regarded the isolate as an example of MGI . Using this method , 85 ( 97 . 7% ) out of the 87 isolates were MGIs . Focusing on each locus , the MGI rate per locus varied from 0 . 0% to 83 . 9% ( average 29 . 1% ) ( Table 1 ) . Focusing on the number of MGI loci per isolate , we found an interesting distribution pattern , similar to an F-distribution ( Fig . 2 ) . In the present study , the highest frequency of MGI loci per isolate was 2 ( found in 25 isolates ) . The frequency decreased gradually , that is , 3 MGI loci; 19 isolates , 4 MGI loci; 12 isolates , 5 MGI loci; 6 isolates , and so on . We suspect that this distribution pattern may vary in each endemic area with different endemicity . In the case of P . falciparum populations , the levels of genetic diversity are normally associated with the levels of malaria endemicity . That is , the levels of genetic diversity of the parasite populations are higher in high transmission areas and lower in low transmission areas [4] , [5] , although some exceptions have been reported [36] . We suspect that there will also be some association between the levels of genetic diversity and the levels of malaria endemicity in P . vivax populations , even though the correlation between these factors is not clearly understood at the time of writing . In P . vivax populations , the levels of genetic diversity tend to be high even in low transmission areas [11]–[13] , [28] . This tendency can likely be attributed to unique biological features of P . vivax , such as early gametocytogenesis and relapse . Early gametocytogenesis may enhance the efficiency of transmission to Anopheles mosquitoes , allowing transmission to occur before symptoms appear – or , more importantly , before antimalarial drugs are administered . Relapses may also enhance the transmission and increase the genetic diversity of P . vivax populations , because the relapse will increase the probability of the coexistence of multiple genotype clones in a single patient , which are subsequently sucked up by an Anopheles mosquito in a single meal . Thus , the levels of genetic diversity of P . vivax populations could be higher than those observed in P . falciparum populations even in low transmission areas . In the present study , the levels of genetic diversity of the South Korean population between 1994 and 2000 ( when the number of malaria cases increased ) were relatively lower than the levels of genetic diversity between 2001 and 2008 ( when the number decreased ) . On the contrary , the levels of multilocus LD in the population between 1994 and 2000 were relatively higher than those between 2001 and 2008 . These results suggested that the latter population was more genetically diverse and had less inbreeding . Furthermore , we divided the population into 3 groups each covering 5-year periods ( 1994–1998 , 1999–2003 and 2004–2008 ) and reexamined the levels of genetic diversity and multilocus LD . Then , we again observed that the levels of genetic diversity in the populations had gradually increased , whereas the levels of multilocus LD had gradually decreased even though there was still strong multilocus LD in the most recent population ( 2004–2008 ) . This result was surprising to us because we expected that the effective population size of the latter population would have decreased due to the reduction in the number of alleles in the population . However , this was not the case . In the South Korean populations , the association between the diversity and the endemicity of the P . vivax population is elusive . There are , however , at least two possible explanations for this result . One is that the levels of genetic diversity of the P . vivax population increased in North Korea from 2001 to 2002 , while the number of vivax malaria cases was very high ( 296 , 540 cases in 2001 , 241 , 190 cases in 2002 ) [1] . Some of the isolates might have then been introduced to South Korea from North Korea by Anopheles mosquitoes . The other possible explanation is that the genetic diversity began accumulating in the South Korean population after the re-emergence in 1993 . If the latter hypothesis is correct then the malaria control program conducted by the South Korean government might not have affected the parasite population structure . One of the clear differences between the P . vivax population in South Korea and populations in tropical and subtropical areas is the pattern of transmission: in South Korea , vivax malaria is seasonally prevalent with a peak during July and August and no transmission in the winter season [37] and very long incubation periods with 8 to 13 months [14] , suggesting that the chance for the recombination of the genome is limited to specific time periods within the year , possibly once or at most twice a year . In fact , we found strong LD in the South Korean population , suggesting that the frequency of recombination in this population would be very limited . However , these results might be associated with the location of the examined MS DNA loci: the 6 loci are in a gene coding a protein and another locus is between 2 genes coding respective proteins . In tropical and subtropical areas , on the other hand , vivax malaria is prevalent throughout the year , and thus recombination may occur throughout the year; this would lead to an increase in the levels of genetic diversity in tropical and subtropical areas . Indeed , in the populations from the Brazilian Amazon , identical haplotypes were rarely observed two years in a row , even in the same endemic area [12] . This would suggest that frequent recombinations occurred between the clones in the population . The present study showed evidence of a low recombination rate and low frequencies of strand-slippage events of the microsatellite loci during mitotic replication in the P . vivax population of South Korea in comparison to populations in tropical and subtropical areas [12] , and demonstrated that the 2 dominant haplotypes ( H16 and H25 ) had been transmitting for several years ( H16; 1996 , 1998–2001 , 2005 , H25; 1995–2001 , 2003 ) ( Table 5 ) . This continuous existence of the same haplotypes for several years is definitive evidence of a low recombination rate in the South Korean P . vivax population . This continuous existence of the same haplotypes could be explained by a local adaptation to vector species . According to Joy et al . [38] , for example , P . vivax in southern Mexico was genetically differentiated into 3 populations . They suggested that this differentiation would be the result of adaptation to different Anopheles species . On the other hand , in South Korea , Anopheles sinensis is a main vector of P . vivax and the other Anopheles species are very minor . Therefore , continuous existence of the predominant haplotypes could not be explained by a local adaptation to certain vector species in this country . There might be some other advantages of these haplotypes , or simply , the variation in the P . vivax population on the Korean peninsula had been very small owing to an effective national eradication program conducted by the National Malaria Eradication Service under the operation of the South Korean government with the support of the WHO in the 1970s [14]–[16] . Although the predominant haplotypes ( H16 and H25 ) and their relatives had been transmitting in the DMZ for a long time , their transmission ended in 2005 . We speculate that these predominant haplotypes were probably eliminated by the malaria control programs conducted by the North Korean government . In fact , according to the WHO World Malaria Report 2010 , the number of vivax malaria cases in North Korea decreased substantially ( 2001: 296540 cases , 2005: 11507 cases ) . The reason for this reduction was not mentioned in detail , however this is probably due to the effect of mass drug administration by the North Korean government supported by South Korea . We suspect that the population structure of P . vivax in North Korea was changed dramatically and that these predominant “old” haplotypes were eliminated completely or became very minor in both the North Korean and South Korean populations . We suspect that the South Korean P . vivax population is a subpopulation of the North . Our previous genetic epidemiological analyses of the South Korean P . vivax population using antigenic molecules [21]–[25] and the mitochondrial genome [39] showed that there were 2 types ( or groups ) of parasite populations in the endemic area . In these previous studies we examined groups of isolates collected from vivax malaria patients in the DMZ in 1997 [21] , 1998 [22]–[24] , 1999 [39] . In the present study , we examined isolates collected from patients in the DMZ between 1994 and 2008 using 10 highly polymorphic microsatellite loci . Once again , we observed two types of parasite populations ( Fig . 4 ) . However , some other haplotypes ( clones ) have been observed in the endemic area since 1998 . The new haplotypes were genetically different to the 2 major groups that have been transmitted since the beginning of the re-emergence ( Fig . 4 ) . This finding was consistent with the results of analyses by Choi et al . using the DNA sequences of 2 antigenic molecules ( circumsporozoite protein , merozoite surface protein-1 ) of isolates collected in the DMZ from 1996 to 2007 [40] . They also reported that new genotypes have been observed since 2000 and that the new genotypes had been rapidly disseminated in the endemic area . The genetic differences between the 2 major groups and the new haplotypes in our data suggested two possibilities: the new haplotypes could have arisen in the DMZ in South Korea through recombination between existing clones in the population; or their emergence could be attributed to a continual introduction of P . vivax from other population sources , probably from North Korea . The present study suggested a low recombination rate in the South Korean population and would seem to indicate that the latter possibility is more likely . A less likely possibility is that all of the isolates examined in this study were continually introduced from North Korea because all the isolates were collected from South Korean soldiers who served in the DMZ . These patients were normally treated by chloroquine within 4 days of the onset of the symptoms , and then treated by primaquine as a radical cure . The recurrence rate ( both new infection and relapse may be included ) of vivax malaria among them is 1 . 6% ( 62 cases of 3881 cases ) and the definitive relapse rate is only 0 . 2% ( 8 cases of 3881 cases ) [41] . In addition , the incubation period of P . vivax on the Korean peninsula is very long ( 8 months to 13 months ) [14] and the transmission is mainly in summer [38] . Moreover , the period of conscription is about 2 years . Therefore , it might have been very difficult to transmit continuously among the South Korean soldiers in the DMZ , leading to the high recombination rate of the genome within the parasite population of the study area . There are a number of sampling limitations to the present study . The number of isolates per year was relatively small ( 2 to 14 isolates , average: 5 . 8 isolates/year ) , and the sample size during the years 2004–2008 was particularly small ( 2 to 7 isolates , average: 3 . 6 isolates/year ) . Moreover , all of the isolates used in this study were collected only from South Korean soldiers or veterans and not from civilians , whose proportion among vivax malaria patients in South Korea has been gradually increasing [18] . In order to overcome these limitations and more accurately estimate the current status of the parasite population in South Korea , it will be necessary to include new isolates collected from civilians in the endemic areas and to increase the sample size of recent years . Although travel between South and North Korea is basically restricted and the malaria control programs in the two countries may not be the same , we suspect that the South Korean P . vivax population is a subpopulation of the North Korean population because the majority of malaria patients live near the border [42] . Anopheles mosquitoes can fly over the DMZ , and South Korean travelers are allowed to visit some parts of North Korea , such as Kaesong and Kumgang-san , which are very famous for sightseeing . Furthermore , from 2001 to 2009 the number of vivax malaria cases in North Korea ranged from twice as high as the number in South Korea to many times higher , indicating that the size of the parasite population in North Korea is probably larger . Thus , the inclusion of North Korean isolates in the analyses would greatly enhance the accuracy of the estimation of the parasite population structure and the transmission dynamics and provide a more complete picture of the P . vivax population in the Korean peninsula; unfortunately the feasibility of doing this is low . In conclusion , molecular epidemiology using highly polymorphic DNA markers of the P . vivax population is a very useful tool for assessing the population structure and transmission dynamics of the parasites , the knowledge of which may lead to the effective control of vivax malaria in the respective endemic areas . | Vivax malaria is widely prevalent , mainly in Asia and South America with 390 million reported cases in 2009 . Worldwide , in the same year , 2 . 85 billion people were at risk . Plasmodium vivax is prevalent not only in tropical and subtropical areas but also in temperate areas where there are no mosquitoes in cold seasons . While most malaria researchers are focusing their studies on the parasite in tropical areas , we examined the characteristics of P . vivax in South Korea ( temperate area ) temporally , using 10 highly polymorphic microsatellite DNA ( a short tandem repeat DNA sequence ) in the parasite genome , and highlighted the differences between the tropical and temperate populations . We found that the South Korean P . vivax population had low genetic diversity and low recombination rates in comparison to tropical P . vivax populations that had been reported . We also found that some of the parasite clones in the population were changing from 1994 to 2008 , evidence suggesting the continual introduction of the parasite from other populations , probably from North Korea . Polymorphic DNA markers of the P . vivax parasite are useful tools for estimating the situation of its transmission in endemic areas . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"ecology",
"epidemiology",
"global",
"health",
"genetics",
"biology",
"population",
"biology",
"public",
"health",
"genetics",
"and",
"genomics"
] | 2012 | Population Structure and Transmission Dynamics of Plasmodium vivax in the Republic of Korea Based on Microsatellite DNA Analysis |
Patients infected by Plasmodium vivax or Plasmodium ovale suffer repeated clinical attacks without primaquine therapy against latent stages in liver . Primaquine causes seriously threatening acute hemolytic anemia in patients having inherited glucose-6-phosphate dehydrogenase ( G6PD ) deficiency . Access to safe primaquine therapy hinges upon the ability to confirm G6PD normal status . CareStart G6PD , a qualitative G6PD rapid diagnostic test ( G6PD RDT ) intended for use at point-of-care in impoverished rural settings where most malaria patients live , was evaluated . This device and the standard qualitative fluorescent spot test ( FST ) were each compared against the quantitative spectrophotometric assay for G6PD activity as the diagnostic gold standard . The assessment occurred at meso-endemic Panenggo Ede in western Sumba Island in eastern Indonesia , where 610 residents provided venous blood . The G6PD RDT and FST qualitative assessments were performed in the field , whereas the quantitative assay was performed in a research laboratory at Jakarta . The median G6PD activity ≥5 U/gHb was 9 . 7 U/gHb and was considered 100% of normal activity . The prevalence of G6PD deficiency by quantitative assessment ( <5 U/gHb ) was 7 . 2% . Applying 30% of normal G6PD activity as the cut-off for qualitative testing , the sensitivity , specificity , positive predictive value , and negative predictive value for G6PD RDT versus FST among males were as follows: 100% , 98 . 7% , 89% , and 100% versus 91 . 7% , 92% , 55% , and 99%; P = 0 . 49 , 0 . 001 , 0 . 004 , and 0 . 24 , respectively . These values among females were: 83% , 92 . 7% , 17% , and 99 . 7% versus 100% , 92% , 18% , and 100%; P = 1 . 0 , 0 . 89 , 1 . 0 and 1 . 0 , respectively . The overall performance of G6PD RDT , especially 100% negative predictive value , demonstrates suitable safety for G6PD screening prior to administering hemolytic drugs like primaquine and many others . Relatively poor diagnostic performance among females due to mosaic G6PD phenotype is an inherent limitation of any current practical screening methodology .
Glucose-6-phosphate dehydrogenase deficiency ( G6PDd ) is the most common inherited disorder , affecting about 400 million people [1–3] . G6PD enzyme catalyzes the first and rate-limiting reaction of the pentose phosphate pathway , the only means of reducing nicotinamide adenine dinucleotide phosphate ( NADPH ) in red blood cell cytosol . In turn , NADPH is the sole source of electrons for reducing glutathione , the principal means of maintaining healthy reduction-oxidation ( redox ) equilibrium in cytosol . Oxidative stress upon red blood cells with impaired G6PD activity leads to threatening redox imbalance . Most people with G6PDd nonetheless lead healthy lives of normal longevity , and it is only exposure to certain drugs , chemicals , foods or infections that impose hemolytic crisis and risk of serious harm . In the malaria endemic rural tropics , the most threatening scenario is becoming infected by the parasite Plasmodium vivax and being prescribed the drug primaquine to prevent the repeated clinical attacks ( called relapses ) deriving from latent liver stages called hypnozoites [2] . Primaquine is an 8-aminoquinoline drug licensed as a hypnozoitocide in 1952 and remains the only therapy available for preventing relapse [4] . During its clinical development in and after World War II , investigators observed hemolytic sensitivity in some subjects . Only later , in 1956 , did those investigators identify deficiency in G6PD as the cause of that sensitivity [5] . Those early studies , conducted in prisoner volunteers in the USA , characterized primaquine sensitivity in African-Americans expressing the A- variant of G6PDd typically expressing 10–20% of normal G6PD activity . Primaquine hemolyzed only older red blood cell populations in those subjects and hemolysis ceased despite continued exposure to high doses of primaquine [6 , 7] . These findings led to the view of primaquine-induced hemolysis as relatively mild and self-limiting . However , studies in the 1960s revealed other variants of G6PDd , including the exquisitely primaquine-sensitive Mediterranean variant [8–12] . Mediterranean variant typically exhibited <5% of normal activity and primaquine-induced hemolysis occurred even among the youngest red blood cells—severe and unlimited hemolysis without cessation of dosing . Over the decades that followed , confirmed severe hemolytic crises and deaths due to primaquine toxicity in G6PD deficient patients accumulated [13–17] . In South and Southeast Asia , where more than 80% of vivax malaria attacks occur , the extraordinary diversity of G6PDd is dominated by Mediterranean-like , severely deficient variants [18 , 19] . Recent recognition of Plasmodium vivax as a pernicious infection and its multiple relapses a serious clinical and public health threat [20–22] elevated awareness of the problem of G6PDd as a very significant barrier to safe primaquine therapy [23 , 24] . Absent the ability to identify G6PDd patients among those infected by P . vivax , providers must choose between risk of harm caused by the drug and that caused by the repeated clinical attacks allowed by withholding the drug . Resolving this therapeutic dilemma requires identifying those at risk of harm with primaquine therapy and thus ensuring those not at risk obtain the enormous therapeutic benefit of primaquine . The most widely used and recommended procedure for screening patients for G6PD deficiency , excluding newborn screening , is the fluorescent spot test ( FST ) , described by G6PD pioneer Ernest Beutler in 1966 [25] . In a laboratory setting the test is relatively simple and inexpensive . However , in the setting of the impoverished rural tropics , the requirements for laboratory skills , refrigeration , specialized equipment , and high costs have excluded its availability to the vast majority of patients suffering malaria . Expert consensus defined practicality criteria for point-of-care G6PD diagnostics that included simplicity of use , ease of interpretation , no specialized equipment or cold chain , and relatively low cost [26–28] . Expert consensus also acknowledged that the availability of such robust devices where most malaria patients live is a key to the control and elimination of endemic P . vivax malaria [28] . In the current study , the performance of the CareStart G6PD ( G6PD RDT , AccessBio , USA ) device against the FST using quantitative spectrophotometric G6PD assay as diagnostic gold standard was compared among residents in a malaria-endemic area of rural eastern Indonesia . The G6PD RDT performed as well as the FST .
This study has been ethically approved by the Eijkman Institute Research Ethics Commission ( EIREC ) ( project No . 69 , February 13th , 2014 ) . After obtaining the informed consent from 610 healthy subjects at least 6 years old , recruitment ceased at targeted full enrollment . Written informed consent was obtained from all study participants . Parents or guardians signed the informed consents for minors under 18 years of age . The village Panenggo Ede is located in the western coastal region of Southwest Sumba regency ( Fig 1 ) , where G6PDd prevalence was known to be >5% [29] . A total of 1117 people resided in this village . Fig 2 shows the work flow where the research team engaged the community gathered at churches or other social functions , and explained the study procedures and intent . Residents were then invited to a study center established in the village at designated times and dates between April and May 2014 . The inclusion criteria were people ≥6 years old , healthy and willing to sign informed consent . A total of 350 females and 260 males provided a 3mL sample of whole venous blood collected into tubes containing EDTA anticoagulant . Samples were held at 4°C prior to processing and analysis on-site ( G6PD RDT ) or nearby temporary laboratory ( FST ) within 3 hours on the same day , or within 3 days at the laboratory in Jakarta ( quantitative G6PD ) . The principle of the CareStart G6PD T screening test is reduction of a colorless nitro-blue tetrazolium dye to purple colored formazan . Thus , whereas a colorless test outcome indicates G6PD deficiency , a purple color reflects G6PD activity ( Fig 3 ) . Readers of the test were instructed to consider only a diagnosis of deficient or normal , with the demand to classify as deficient any test strip exhibiting a colorless to distinctly lighter hue of purple compared to that of most other tests . This approach would ensure safety when primaquine therapy would follow the diagnosis of G6PD normal . Two microliters of whole blood was removed from the EDTA tube by a stick device included in the RDT kit and placed into the sample window , immediately followed by two drops of a provided buffer solution into the assay window according to the manufacturer’s instructions . After ten minutes at the ambient temperature of approximately 30°C , the RDT was visually read and classified as deficient or normal . At the end of each day of work in the village , venous blood was transferred on ice packs to a field laboratory in Weetabula to conduct the fluorescent spot test ( FST , Trinity Biotech , Ireland; Cat . No . 203-A ) using deficient ( Cat . No . G5888 ) , intermediate ( Cat . No . G5029 ) and normal ( Cat . No . G6888 ) G6PD controls from the same company . This qualitative test is a modification of Beutler’s test in which glucose-6-phosphate and NADP+ reagents ( substrate solution ) in the presence of G6PD sample produce fluorescent NADPH and 6-phosphogluconate . Progress of the reaction was observed in the dark under long-wave ultraviolet illumination of sample filter paper ( Whatman No . 1 filter paper , Cat . No . 1001–150 ) at intervals of zero , 5 and 10 minutes . Briefly , 200 μl of substrate solution and 10 μl of gently mixed venous blood was put into a 5 ml tube and mixed by manual swirling . A single drop of this solution was transferred onto filter paper marked as time zero . The tube was then placed into a 37°C water bath for 5 min , when another drop was placed onto filter paper marked as time 5 min . This was repeated for the final sample at 10 min . The filter papers were allowed to dry at room temperature ( 25°-29°C ) before visual inspection under UV light in an otherwise dark room . Deficient ( no fluorescence ) , intermediate ( weak fluorescence ) and normal ( strong fluorescence ) controls were done for every set of 10 samples from the subjects . Readers were instructed to classify intermediate test outcomes as deficient . The principle of the G6PD quantitative assay from Trinity Biotech ( Cat . No . 345-B ) is similar to the FST . Fluorescence from NADPH produced in the same substrate solution mixture was read at 340nm using a high-grade , temperature-controlled spectrophotometer ( UV-1800 UV-VIS Shimadzu ) . The same G6PD controls from Trinity Biotech were conducted for each set of 25 samples from the subjects . The assay was performed in an air-conditioned ( ~25°C ) laboratory at the Eijkman Institute in Jakarta within 3 days of blood withdrawal . The venous blood tubes were kept at 4°C at all times prior to use in Jakarta . Hemoglobin level was determined using 10 μl of blood into a micro-cuvette supplied by the manufacturer of the HemoCue system ( HemoCue AB , Sweden ) and immediately read in the instrument ( Hb201+ ) of that system for hemoglobin measurement prior to the G6PD quantitative assay . The manufacturer’s instructions were strictly followed for measuring absorbance at 340nm and deriving an estimate of G6PD activity in U/g Hb at 30°C using the incubated spectrophotometer . Although the manufacturer recommended a cut off value <4 . 6 U/g Hb for deficient activity , we selected <5U/g Hb on the basis of prior survey in the same area showing a median G6PD activity of 10 U/g Hb [29] . We aimed for a 50% cut off value , that being the limit of relative safety with respect to potential hemolytic loss of red blood cell populations , knowing this value correlated with the proportion of deficient red blood cells in a laboratory model of the female heterozygous state [30] . The assay was performed in triplicates where a mean was derived to be used for downstream analyses . Samples for G6PD genotyping were selected on the basis of a deficient classification by G6PD quantitative assay ( <5U/gHb ) , or by having Hb <8g/dL ( Fig 2 ) . DNA from the buffy coat of venous blood in EDTA tubes was extracted using QIAamp DNA Blood Mini Kit ( Qiagen , Cat . No . 51106 ) . DNA from subjects classified as deficient by quantitative G6PD assay was examined by PCR/RFLP for the most common variants in Sumba: Vanua Lava , Viangchan , Chatham , and Kaiping . Table 1 details the primers employed and the PCR and RFLP products thus expected . PCR conditions were as follow: 1X buffer GC Hifi ( Kapa Biosystem ) , 200 μM dNTPs , 200 μM forward and reverse primer each , 0 . 4 U Kapa Hifi polymerase in 25 μl PCR reaction . PCR cycle for the variants were also the same except in the annealing temperature: 95°C 5 min before entering PCR cycle of 30X; denaturation at 95°C for 30 sec; and annealing 65°C , 56°C , 61°C and 62°C , for Vanua Lava , Viangchan , Chatham and Kaiping respectively . Each was followed by extension at 72°C for 30 sec , where another 7 min at 72°C was needed at the end of the 30 cycles . The PCR products were cut with the restriction enzymes as listed in Table 1 . After incubation at 37°C overnight , products were run on 3% agarose gel for analysis . The samples testing as normal for the common variants were PCR and whole-gene sequenced as described by others [31] . Sequences were aligned to G6PD reference sequence from NCBI , NG_009015 . 2 . DNA extracted from venous blood was also genotyped for Southeast Asian ovalocytosis ( SAO ) , alpha thalassemia , and hemoglobin E ( HbE ) . Table 1 lists the primers for SAO , one and two gene deletions for alpha thalassemia and HbE . The PCR conditions for each mutation were as reported elsewhere [32] . PCR conditions for one-gene deletions were as previously reported [33] . In the current study the diagnostic objective was not G6PD deficiency per se , but a diagnostic outcome indicating either hazard or safety with administration of a potentially hemolytic drug . As such , diagnostic performance of the G6PD screening techniques was linked to the perceived primaquine safety margin of 30% of normal activity per WHO recommendation [26] . We aimed to classify all male hemizygotes and female heterozygotes having less than variable thresholds of normal G6PD activity ( <10% , <30% , or <60% ) as deficient . The median G6PD activity among subjects having ≥5U/g Hb was considered 100% of normal . These thresholds represented an examination of variance in diagnostic performance representing poor , good , and complete safety , respectively , with respect to exposure to primaquine . Poor safety at 10% would likely include patients at risk of hemolysis , whereas complete safety at a 60% would unnecessarily deny some patients primaquine treatment . The 30% cut off value represents a compromising balance of those problems . Diagnostic performance of the qualitative G6PD RDT and FST were assessed against the quantitative G6PD classification as “deficient” or “normal” at G6PD activity thresholds . Further , the analyses were segregated by sex for the simple reason that hemizygosity versus heterozygosity ( males and females , respectively ) profoundly impacts diagnostic performance for G6PD deficiency [34] . Males tend to be wholly deficient or normal , whereas females will present the full spectrum of G6PD activity levels due to mosaicism of this X-linked trait [35] . Standard methods for calculation of sensitivity , specificity , positive predictive value , and negative predictive value were applied to the G6PD RDT and FST for each threshold of percent of normal G6PD activity . The meaning of these parameters in the context of a diagnosis guiding primaquine therapy bears explanation here . Sensitivity and specificity are easily grasped , i . e . , rate of true positives and rate of true negatives , respectively . The terms “positive” and “negative” refer to what is defined here as “deficient” and “normal” G6PD phenotype , respectively . A test negative for G6PD activity is positive for G6PD deficiency , and vice versa for a positive test for G6PD activity . The terminology “deficient” ( positive for deficiency ) versus “normal” ( negative for deficiency ) recommended by WHO [26] , avoids confusion and was adopted here . Further , the terms “deficient predictive value” ( DPV ) and “normal predictive value” ( NPV ) were applied for consistency and clarity , but using precisely the same standard mathematical methods for all of these statistics . DPV estimates the probability that those classified as deficient truly are , whereas NPV estimates the probability that those classified as normal truly are . In the primaquine therapy context of G6PD diagnostics , the most important statistic is NPV because a diagnosis of normal prompts exposure to primaquine . Fig 4 illustrates the rationale at work . In a practical sense , NPV estimates the probability of primaquine being safely administered , whereas DPV reflects the proportion of patients being denied primaquine therapy who actually cannot take it safely . Statistical significance of diagnostic performance indicator by diagnostic test was evaluated by Chi-square test . Sensitivity , specificity , DPV and NPV were presented using proportion analysis and Fisher’s exact 95% confidence intervals . Mean and range of hemoglobin level were calculated to determine distribution by gender . Data were analyzed using Stata 9 .
The overall prevalence of G6PDd at Panenggo Ede by quantitative assay ( <5U/g Hb ) was 7 . 2% ( 44/610 ) , 9 . 2% for males ( 24/260 ) and 5 . 7% ( 20/350 ) for females . Southeast Asian ovalocytosis ( SAO ) occurred in 12 . 7% ( 78/610 ) , alpha thalassemia ( alpha Thal ) in 15 . 1% ( 92/610 ) , and hemoglobin E ( HbE ) in 16 . 4% ( 100/610 ) of residents . Double mutations occurred among 13 residents having G6PDd ( 5 with SAO , 3 with alphaThal , 5 with HbE ) , and one subject had G6PDd , SAO and HbE . SAO occurred in 21 subjects also having alpha Thal , and in 7 people also having HbE . In total , 44 . 3% ( 270/610 ) of the population had one or more of these four blood disorders . Table 2 summarizes findings of malaria and anemia in the community . The overall prevalence of microscopically patent parasitemia was 2 . 5% ( 15/610 ) ; with 53% P . falciparum , 33% P . vivax , and 14% mixed by these species . The mean level ( and range ) of Hb in the study population was 13 . 2 ( 6 . 0–22 . 8 ) g/dL . Only 3 subjects had levels <8 . 0g/dL , and the majority had ≥10 . 0g/dL ( 607; 99 . 5% ) . Males and females had similar but statistically distinct levels of Hb: 13 . 8 ( 6 . 9–20 . 2 ) , and 12 . 8 ( 6 . 0–22 . 8 ) , respectively ( P<0 . 0001 ) . Among the three severely anemic subjects ( <8 . 0g/dL ) , genotyping for G6PD variants revealed one as a female ( Hb 6 . 0g/dL ) heterozygous for Vanua Lava variant with a quantitative G6PD value of 13 . 55 U/gHb . The other two were G6PD normal genotype and phenotype . Hemoglobin level did not appear to be significantly different between subjects with or without any particular inherited blood disorder evaluated . Fig 5A illustrates the results of genotyping of the 44 subjects deemed G6PDd by quantitative assay ( <5 . 0 U/g Hb ) . Vanua Lava dominated at 50% ( 22/44 ) , followed by Viangchan at 30% ( 13/44 ) , Coimbra Shunde at 11% ( 5/44 ) , Chatham at 7% ( 3/44 ) , and 1 subject was not successfully genotyped ( 2% ) . Fig 5B illustrates G6PD activity values for subjects classified as normal by G6PD activity , as well as with those classified as deficient and successfully genotyped . Heterozygous females having ≥5 . 0 U/gHb would have been excluded from the genotyping survey and would be included among normals in the figure . The values illustrated for heterozygotes inform only the diagnostic assessment rather than as a survey of their G6PD activity ranges . Among hemizygous males , however , the G6PD activity mean and range may be considered estimates of residual enzyme activity among the specific variants: 0 . 8 U/g Hb ( 0 . 27–2 . 5 U/g Hb ) for Vanua Lava; 0 . 97 U/g Hb ( 0 . 52–1 . 62 U/g Hb ) for Viangchan; and 0 . 09 U/g Hb ( 0 . 03–0 . 16 U/g Hb ) for Coimbra Shunde . Remarkably low G6PD activity was also observed in the two females expressing Coimbra Shunde variant ( 0 . 57 U/g Hb; the mean of 0 . 11 and 1 . 04 U/g Hb ) . Chatham variant was found only in 3 females . G6PD activity did not vary with age in this study , as reported in another study from the same region [29] . Fig 6 illustrates diagnostic outcomes for the G6PD RDT and FST across quantitative G6PD activity values in males and females . The tests performed similarly , with each discerning G6PD deficiency at a threshold of 10% of normal activity among males and females . However , 2 FST tests in males were read as normal at <10% of activity . The tests also performed similarly at a 60% activity threshold for both tests . The FST in males showed a propensity for false deficient reads , even at or above 100% of normal activity , but was especially frequent between 65% and 85% of normal activity . Three false deficient reads occurred among males with the G6PD RDT at 65% , 90% , and 115% of normal G6PD activity . Although both tests properly identified all female heterozygotes below a 30% threshold ( with a single exception for the G6PD RDT at 22% of normal activity ) , each also exhibited a profound propensity for false deficient reads all across the range of G6PD activity values . Table 3 summarizes the statistical analyses of these diagnostic outcomes among males and females diagnostic thresholds of 10% , 30% , and 60% for both of the qualitative G6PD screening kits . At the 30% threshold the G6PD RDT showed superior sensitivity and specificity in males compared to the same for the FST: 100% and 98 . 7% versus 91 . 7% and 92 . 4% , respectively ( P = 0 . 48 and P < 0 . 001 for sensitivity and specificity respectively ) . Among females at the 30% threshold , no statistically significant differences occurred between the sensitivities and specificities of the two kits: 83 . 3% and 100% vs . 92 . 7% and 92 . 2% ( P = 1 and P = 0 . 89 ) for G6PD RDT vs . FST , respectively . Deficient predictive value ( DPV ) for the G6PD RDT for males at 30% of normal G6PD activity was superior to the same with FST: 63 . 0% versus 37 . 5% ( P = 0 . 05 ) , respectively . Among females at the same threshold , DPV was 10 . 0% and 9 . 1% ( P = 1 ) . Among males for both G6PD RDT and FST at 30% threshold , normal predictive value ( NPV ) was 100% and 99 . 1% respectively ( P = 0 . 23 ) , and for females 100% and 100% ( P = 1 ) .
This study affirms the good diagnostic performance of a new qualitative G6PD screening device , the G6PD RDT , intended for use at the point-of-care typical of where most malaria patients live . The G6PD RDT always correctly classified male patients with severe G6PD deficiency , whereas the FST failed to do on two occasions—a serious problem imposing risk of harm with primaquine therapy . Both screening kits often misclassified G6PD normal subjects as deficient , which would result in withholding primaquine therapy from patients who could safely consume it . All qualitative tests for G6PD suffer the drawback of classifying many female heterozygotes as G6PD normal despite significantly impaired G6PD activity ( i . e . , 30% to 60% of normal ) , exposing them to risk of harm with primaquine therapy . The degree of that risk is poorly understood and requires a great deal more work , both in terms of assessing it and mitigating it with improved diagnostics . There is also a need to evaluate the stability of the RDT during storage in the field . | G6PD is an enzyme that chemically protects us from otherwise toxic substances , like some chemotherapeutic agents . About 8% of people exposed to malaria have an inherited disorder that impairs G6PD activity , leaving them vulnerable to harm by an important therapy against malaria , primaquine . This drug alone prevents repeated clinical attacks stemming from dormant parasites residing in the human liver . Absent certain knowledge of patient G6PD status , healthcare providers managing patients infected by Plasmodium vivax or Plasmodium ovale malaria must choose between risk of harm caused by hemolytic toxicity of primaquine and that caused by the parasite after withholding therapy . Resolving that therapeutic dilemma requires assessment of patient G6PD status at the point-of-care in the impoverished rural tropics , where the vast majority of malaria patients live . Current technology for such screening is impractical in that setting . In this study we evaluated screening designed for practicality at the endemic tropical point-of-care: a rapid diagnostic test for G6PD ( G6PD RDT; CareStart G6PD , AccessBio , USA ) . We found the G6PD RDT to be effective in screening volunteers living in rural eastern Indonesia . This G6PD RDT kit costs relatively little ( $1 . 50 ) , was simple to execute and interpret , required no specialized equipment or skills , performed well at ambient tropical temperatures ( >30°C ) , and required no cold chain storage . This and similar kits may permit safe universal access to primaquine therapy against relapse of P . vivax , a vitally important step forward in mitigating the global burden of morbidity and mortality imposed by this pernicious parasite . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"parasite",
"groups",
"body",
"fluids",
"plasmodium",
"drugs",
"tropical",
"diseases",
"parasitic",
"diseases",
"anemia",
"parasitic",
"protozoans",
"parasitology",
"organisms",
"apicomplexa",
"protozoans",
"red... | 2016 | Assessment of Point-of-Care Diagnostics for G6PD Deficiency in Malaria Endemic Rural Eastern Indonesia |
The blood-feeding hookworm Necator americanus infects hundreds of millions of people worldwide . In order to elucidate fundamental molecular biological aspects of this hookworm , the transcriptome of the adult stage of Necator americanus was explored using next-generation sequencing and bioinformatic analyses . A total of 19 , 997 contigs were assembled from the sequence data; 6 , 771 of these contigs had known orthologues in the free-living nematode Caenorhabditis elegans , and most of them encoded proteins with WD40 repeats ( 10 . 6% ) , proteinase inhibitors ( 7 . 8% ) or calcium-binding EF-hand proteins ( 6 . 7% ) . Bioinformatic analyses inferred that the C . elegans homologues are involved mainly in biological pathways linked to ribosome biogenesis ( 70% ) , oxidative phosphorylation ( 63% ) and/or proteases ( 60% ) ; most of these molecules were predicted to be involved in more than one biological pathway . Comparative analyses of the transcriptomes of N . americanus and the canine hookworm , Ancylostoma caninum , revealed qualitative and quantitative differences . For instance , proteinase inhibitors were inferred to be highly represented in the former species , whereas SCP/Tpx-1/Ag5/PR-1/Sc7 proteins ( = SCP/TAPS or Ancylostoma-secreted proteins ) were predominant in the latter . In N . americanus , essential molecules were predicted using a combination of orthology mapping and functional data available for C . elegans . Further analyses allowed the prioritization of 18 predicted drug targets which did not have homologues in the human host . These candidate targets were inferred to be linked to mitochondrial ( e . g . , processing proteins ) or amino acid metabolism ( e . g . , asparagine t-RNA synthetase ) . This study has provided detailed insights into the transcriptome of the adult stage of N . americanus and examines similarities and differences between this species and A . caninum . Future efforts should focus on comparative transcriptomic and proteomic investigations of the other predominant human hookworm , A . duodenale , for both fundamental and applied purposes , including the prevalidation of anti-hookworm drug targets .
Soil-transmitted helminths ( = geohelminths ) are responsible for neglected tropical diseases ( NTDs ) mostly in developing countries [1] . In particular , the blood-feeding hookworms Necator americanus and Ancylostoma duodenale ( Nematoda ) infect ∼740 million people in rural areas of the tropics and subtropics [2] , causing an estimated disease burden of 22 million disability-adjusted life years ( DALYs ) [3] . Geographically , N . americanus is the most widely distributed hookworm of humans globally [4] . The life cycle is direct , with thin-shelled eggs passed in the faeces from the infected host . Under suitable environmental conditions ( e . g . , 26°C and 100% humidity; [5] ) , the eggs hatch and develop through two free-living larval stages to the infective , third-stage ( L3; filariform ) larvae . The latter larvae penetrate human skin and migrate via the circulatory system and lung to finally reside as adults usually in the duodenum . The adult stages attach by their buccal capsule to the intestinal mucosa , rupture capillaries and feed on blood . The pathogenesis of hookworm disease is mainly a consequence of the blood loss , which occurs during attachment and feeding . The disease ( = necatoriasis ) is commonly characterized by iron-deficiency anaemia , which can cause physical and mental retardation and sometimes deaths in children , adverse maternal-foetal outcomes [6]–[7] and , in chronically infected individuals , can result in a significant alteration of their immune response to helminths [8] . Traditionally , the control of hookworm disease has relied mostly on the treatment of infected individuals with anthelmintics , such as albendazole , mebendazole , pyrantel pamoate and/or levamisole . With mass treatment strategies now in place in a number of countries [9]–[10] , there is an increased potential for hookworms to develop genetic resistance against the compounds administered , if they are used excessively and at suboptimal dosages . Thus , given the experience with drug resistance in parasitic nematodes of livestock [11] , it is prudent to maintain a continual focus on the discovery of novel drugs against hookworms of humans . Such a discovery effort could be underpinned by an integrated genomic-bioinformatic approach , using functional genomic and phenomic information available for the free-living nematode Caenorhabditis elegans ( see WormBase; www . wormbase . org ) . This nematode , which is the best characterized metazoan organism [12]–[13] , is considered to be relatively closely related to nematodes of the order Strongylida ( to which hookworms belong ) [14] . Current evidence indicates that ∼60% of genes in strongylids ( or hookworms ) have orthologues/homologues in C . elegans [15]–[16] , and that a range of biological pathways is conserved between strongylid nematodes/hookworms and this free-living nematode [17]–[20] . Therefore , conducting comparative explorations of molecular data sets between these nematodes should identify nematode-specific biological pathways , which , if essential for the development and survival , could provide new targets for nematocidal drugs . Next generation sequencing technologies , such as ABI-SOLiD , Illumina/Solexa ( www . illumina . com; [21] ) , Helicos ( www . helicosbio . com; [22] ) and 454/Roche ( www . 454 . com; [23] ) , together with the recent progress in bioinformatics , are providing unique opportunities for the high-throughput transcriptomic and genomic explorations of nematodes in far more detail than previously possible [24] and at a substantially lower cost than using conventional ( Sanger ) sequencing . To date , genomic and molecular studies of hookworms have mainly involved the canine hookworm , Ancylostoma caninum [19] , [25]–[27] , because of its use as a model for human hookworms [27]–[28] . In contrast , genomic datasets for N . americanus are scant , representing a major constraint to progress in molecular research of this nematode [4] . In the present study , we ( i ) conducted a detailed exploration and functional annotation of the transcriptome of the adult stage of N . americanus by 454 sequencing coupled to semi-automated bioinformatic analyses , ( ii ) compared the transcriptome of N . americanus to currently available transcriptomic data for A . caninum , and ( iii ) inferred the essentiality of key genes and gene products in order to predict putative drug targets .
The nucleotide sequence data produced for this study are available in the GenBank database under accession SRA012052 . The contigs assembled from these data can be requested from the primary author or are available at www . nematode . net . The “Shanghai strain” of N . americanus ( kindly provided by Drs Bin Zhan and Peter Hotez ) was produced in golden hamsters ( Mesocricetus auratus; infected for 94 days ) at the Universidade Federal de Minas Gerais , Brazil . The infection experiment was conducted according the animal ethics guidelines of the Universidade Federal de Minas Gerais . Total RNA from 30 adult worms was prepared using TRIzol Reagent ( GibcoBRL , Life Technologies , USA ) following the manufacturer's instructions and then treated with Ambion Turbo DNase ( Ambion/Applied Biosystems , Austin , TX ) . The integrity of the RNA was verified using the Bioanalyzer 2100 ( Agilent Technologies , USA ) , and the yield determined using the NanoDrop ND-1000 UV-VIS spectrophotometer v . 3 . 2 . 1 ( NanoDrop Technologies , Wilmington , DE ) . The cDNA library was constructed using the SMART™ kit ( Clontech/Takara Bio , CA ) from ∼100ng of total RNA . An optimized PCR cycling protocol ( over 20 cycles ) was used to amplify full-length cDNAs , employing primers complementary to the SMART IIA-Probe and custom oligo ( dT ) , and the Advantage-HF 2 polymerase mix ( Clontech/Takara ) . The cDNA was normalized by denaturation-reassociation , treated with duplex-specific nuclease ( Trimmer kit , Evrogen , CA ) and amplified over 11 cycles . Subsequently , the 5′- and 3′- adaptors were removed by digestion with the exonuclease Mme1 and streptavidin-coated paramagnetic beads [29] . The normalized cDNA ( 500–700 bases ) was then amplified using 9 cycles of Long and Accurate ( LA ) -PCR [30] and then sequenced in a Genome Sequencer™ ( GS ) Titanium FLX instrument ( Roche Diagnostics ) employing a standard protocol [23] . Expressed sequence tags ( ESTs ) generated from the normalised cDNA library for N . americanus were assembled and annotated using a standard bioinformatic pipeline [31] . Briefly , sequences were aligned and assembled using the Contig Assembly Program v . 3 ( CAP3; [32] , employing a minimum sequence overlap length of 50 nucleotides and an identity threshold of 95% . ESTs ( n = 2 , 200; www . ncbi . nlm . nih . gov ) from adult N . americanus available from previous studies [4] , [16] , [33] , [34] were included for comparative analysis . Following the pre-processing of the ESTs , contigs and singletons from the present dataset were subjected to analysis by BLASTx ( NCBI , www . ncbi . nlm . nih . gov ) and BLASTn ( EMBL-EBI Parasite Genome Blast Server , www . ebi . ac . uk ) to identify putative homologues in C . elegans , other nematodes , and organisms other than nematodes ( e-value of ≤1e-05 ) . WormBase Release WS200 ( www . wormbase . org ) was interrogated extensively for relevant information on C . elegans orthologues/homologues , including transcriptomic , proteomic , RNAi phenotypic and interactomic data . Gene ontology ( GO ) annotations were conducted using BLAST2GO [35] . Peptides were mapped by InterProScan [36] and linked to respective pathways in C . elegans using the KEGG Orthology-Based Annotation System ( KOBAS , [37] ) . The protein sequences inferred from open reading frames ( ORFs ) of the ESTs with orthologues in C . elegans were also subjected to “secretome analysis” using the program SignalP v . 2 . 0 ( available at www . cbs . dtu . dk/services/SignalP/ ) , employing both the neural network and hidden Markov models to predict signal peptides and/or anchors [38]–[40] . Also , transmembrane domains were inferred using the program TMHMM ( www . cbs . dtu . dk/services/TMHMM/; [41]–[43] ) . Protein sequences inferred from contigs for N . americanus were compared with those predicted for C . elegans and from a similar-sized , publicly available EST dataset for adult A . caninum produced by 454 sequencing ( GenBank accession numbers EW741128-EW744730; EX534506-EX567272 ) ; protein similarities were displayed using SimiTri [44] . All protein sequences predicted from contigs for N . americanus were compared with protein sequences available in the OrthoMCL 2 . 0 database ( www . OrthoMCL . org ) by BLASTp ( e-value cut off of <1e-05 ) . A subset of C . elegans protein homologues was then selected based on: ( i ) an association with a lethal RNAi phenotype; ( ii ) the presence/absence of gene paralogues ( based on OrthoMCL orthology grouping ) ; and ( iii ) GO annotation to terms linked to enzyme or G protein-coupled receptor ( GPCR ) activity ( i . e . , GO:0003824 or GO:0004930 , or a sub-term thereof ) . The following information was obtained: ( i ) network connectivity score ( cf . http://www . functionalnet . org/wormnet/Wormnet_v1_index . html; see [45] ) ; ( ii ) presence of mammalian orthologues ( based on OrthoMCL orthology grouping ( iii ) essentiality information ( i . e . association with non-wildtype RNAi phenotypes ) in other model organisms ( including Saccharomyces cerevisiae , Mus musculus and Drosophila melanogaster ) based on OrthoMCL groups . Each predicted drug target was selected based on ( i ) the presence of orthologues linked to non-wildtype RNAi or mutant phenotypes in S . cerevisiae , M . musculus and D . melanogaster , ( ii ) the absence of orthologues/homologues from the human host and ( iii ) its network connectivity score [45] . To predict the potential of selected C . elegans orthologues of N . americanus contigs as drug targets ( = “druggability” ) , the InterPro domains inferred from the predicted proteins were compared with those linked to known small molecular drugs which follow the ‘Lipinsky rule of 5’ regarding bioavailability [46] , [47] . Similarly , GO terms inferred from the predicted proteins were mapped to Enzyme Commission ( EC ) numbers , and a list of enzyme-targeting drugs was compiled based on data available in the BRENDA database ( www . brenda-enzymes . info; [48] , [49] ) . The C . elegans orthologues included in the list were ranked according to the ‘severity’ of the non-wild-type RNAi phenotypes ( i . e . adult lethal , embryonic and/or larval lethal , sterile and other defects ) in C . elegans ( cf . www . wormbase . org ) defined in previous studies [50] , [51] .
A total of 116 , 839 ESTs ( 287±235 bases in length ) was generated by 454 sequencing . After removing the ESTs of <100 bases , 63 , 523 ESTs were assembled into 19 , 997 contigs ( 369 bases±215 . 31 ) . Of these , 6 , 771 ( 33 . 9% ) had known C . elegans orthologues , and 2 , 287 ( 11 . 4% ) matched known nucleotide sequences from various nematodes , including Brugia malayi , Haemonchus contortus , Pristionchus pacificus , N . americanus , A . caninum , A . duodenale and Nippostrongylus braziliensis ( 73 . 2% ) , other invertebrates ( 21 . 3% ) and some vertebrates ( 5 . 5% ) available in current databases . All of the previously published ESTs for N . americanus ( www . ncbi . nlm . nih . gov; [4] , [16] , [33] , [34] ) represented a subset ( 12 . 4% ) of the present dataset ( not shown ) . The number of ORFs in the N . americanus EST data , predicted peptides and their signal , transmembrane and/or InterPro domains as well as the results of GO and KOBAS ( pathway mapping ) searches are given in Table 1 . A total of 12 , 799 proteins were predicted from the 19 , 997 contigs , of which 7 , 214 mapped to known proteins defined by 2 , 381 different domains ( Tables 1 and S1 ) , the most abundant being ‘WD40’ ( IPR0011680; 10 . 6% ) , ‘proteinase inhibitors’ ( IPR000215; 7 . 8% ) and ‘EF-hand’ molecules ( IPR018248; 6 . 7% ) ( Table 2 ) . The subsequent annotation of the inferred proteins revealed 887 different GO terms , of which 314 were ‘biological process’ , 117 ‘cellular component’ and 456 ‘molecular function’ ( Tables 3 and S2 ) . The predominant terms were ‘translation’ ( GO:0006412 , 20 . 3% ) and ‘metabolic process’ ( GO:0008152 , 14 . 9% ) for ‘biological process’; ‘intracellular’ ( GO:0005622 , 25 . 1% ) and ‘ribosome’ ( GO:0005840 , 17% ) for ‘cellular component’ , and , ‘ATP binding’ ( GO:0005524 , 18 . 9% ) and ‘structural constituent of ribosome’ ( GO:0003735 , 17 . 9% ) for ‘molecular function’ ( Tables 3 and S2 ) . Proteins inferred from the N . americanus contigs were predicted to be involved in 235 different biological pathways , of which the vast majority represented ‘ribosome biogenesis’ ( n = 163 , 70% ) , ‘oxidative phosphorylation’ ( n = 148 , 63% ) and ‘proteases’ ( n = 140 , 60% ) ( see Table S3 ) . For comparative analyses , publicly available EST data for the adult stage of A . caninum was included . For this dataset , the same bioinformatic analyses described in the Methods section were conducted . From 15 , 755 contigs of A . caninum , a total of 12 , 622 proteins were inferred , of which 4 , 534 matched those encoded by N . americanus ORFs ( Figure 1 ) ; 8 , 650 of these predicted proteins could be mapped to known molecules with 2 , 546 different motifs ( Tables 1 and S1 ) . The protein motifs ‘SCP-like extracellular’ ( IPR014044 , 9 . 5% ) , ‘ankyrin’ ( IPR002110 , 7% ) and ‘allergen V5/Tpx-1 related’ ( IPR0011283 , 6% ) were most commonly recorded in the A . caninum dataset ( Table 2 ) . Differences in the numbers of IPR domains identified in the N . americanus and A . caninum predicted peptides were calculated using a Chi-square test ( p<0 . 05 ) and are indicated in Table 2 . GO annotation of the A . caninum predicted peptides revealed 323 different terms for ‘biological process’ , 119 for ‘cellular component’ and 500 for ‘molecular function’ ( Tables 3 and S2 ) . The terms ‘metabolic process’ ( GO:0008152 , 7 . 4% ) and ‘proteolysis’ ( GO:0006508 , 6 . 6% ) had the highest representation for ‘biological process’ , as did ‘intracellular’ ( GO:0005622 , 7 . 5% ) and ‘membrane’ ( GO:0016020 , 6 . 6% ) for ‘cellular component’; and , ‘ATP binding’ ( GO:0005524 , 13 . 4% ) and ‘catalytic activity’ ( GO:0003824 , 9% ) for ‘molecular function’ ( Tables 3 and S2 ) . Using the protein data , a total of 235 different biological pathways were predicted , of which ‘proteases’ ( n = 219 , 93% ) , ‘other enzymes’ ( n = 164 , 70% ) and ‘protein kinases’ ( n = 151 , 54% ) were the most predominant ( see Table S3 ) . From the N . americanus dataset , 5 , 498 proteins matched known proteins encoded by orthologues available in the OrthoMCL 2 . 0 database ( www . OrthoMCL . org ) ; 372 of these proteins had homologues in C . elegans , and 278 ( 277 enzymes and one G-PCR ) of them were linked to adult lethal , embryonic and/or larval lethal and sterile RNAi phenotypes ( Table S4 ) . A subset of 18 molecules in N . americanus with homologues in C . elegans but not in humans were defined , also considering RNAi phenotype/s [i . e . adult lethal ( n = 2 ) , larval and/or embryonic lethal ( n = 16 ) , sterile ( n = 4 ) and other defects ( n = 12 ) ; cf . Table 4] , as drug target candidates . These proteins could be mapped to 54 ‘druggable’ InterPro domains , and 212 EC numbers were linked to ‘druggable’ enzymes; a total of 3 , 320 effective drugs were predicted ( Table S4 ) .
Next-generation sequencing and integrated bioinformatic analyses have provided detailed and biologically relevant insights into the transcriptome of the adult stage of N . americanus . A total of 12 , 799 ORFs were inferred from the present EST dataset , thus increasing the number of predicted proteins currently available ( for this stage/species ) in public databases by approximately 27-fold [4] . Amongst the InterPro domains identified , ‘WD40’ , ‘proteinase inhibitors’ and ‘EF-hand’ motifs were the most abundant , followed by ‘proteases’ and ‘protein kinases’ . WD40 repeats ( also known as WD or beta-transducin repeats ) are short ( ∼40 amino acid ) motifs found in the proteomes of all eukaryotes and implicated in a variety of functions , ranging from signal transduction and transcription regulation to cell-cycle control and apoptosis [52] , [53] . WD40 motifs act as sites for protein-protein interactions; proteins containing WD40 repeats are known to serve as platforms for the assembly of protein complexes or mediators of a transient interplay with other proteins , such as the ubiquitin ligases , involved in the onset of the anaphase during cell mitosis [54] . Similarly , proteins containing ‘EF-hand’ domains are involved in a number of protein-protein interactions regulated by various specialized systems ( e . g . , Golgi system , voltage-dependent calcium channels and calcium transporters ) for the uptake and release of calcium , which acts as a secondary messenger for their activation [55] . In C . elegans , both EF-hand and WD40 proteins are known to be required for the maturation of the nervous system and the formation of ciliated sensory neurons , in particular of the chemoreceptors located in the amphids [8] , [56] . The amphids of parasitic nematodes are , besides having the chemoreceptive activity , also known to play a role as secretory organs , primarily to provide an appropriate substrate for the transmission of neuronal potentials [57] . However , in N . americanus , a group of specialized amphidial neuronal cells ( = amphidial glands; [57] ) expresses a group of aspartic proteases ( i . e . cathepsin D-like Na-APR-1 and Na-APR-2 ) which are proposed to degrade host haemoglobin and serum proteins in the buccal capsule of adult worms [58] . In the dog hookworm , A . caninum , the amphidial glands have also been shown to produce a proteinase inhibitor ( called ‘ancylostomatin’ ) that acts as an anticoagulant to promote the flow of host blood and tissue fluids into the buccal capsule and the intestine of the parasite [59] . Although proteinase inhibitors , such as the ‘kunitz-type’ molecules , were significantly more abundant in the transcriptome of adult N . americanus [4] , [33] than Ancylostoma spp . , they have been better characterized in the latter parasites [60]–[63] for which both single and multiple kunitz-domain proteins have been described [61] . For instance , a cDNA coding a single kunitz-domain proteinase inhibitor ( named AceKI-1 ) was isolated from A . ceylanicum . The corresponding recombinant protein has been shown to act as a tight-binding inhibitor of the serine proteases chymotrypsin , pancreatic elastase , neutrophil elastase and trypsin [60] and confers partial protection against hookworm-associated growth delay in hamsters [62] . Recently , a kunitz-type cDNA was shown to be enriched in the adult male of A . braziliense [63] . Although their precise biological function remains to be determined , kunitz-type proteinase inhibitors of hookworms appear to play pivotal roles in preventing homeostasis and inhibiting host proteases ( e . g . , pancreatic and intestinal enzymes; [60] , [64] ) . Proteases were also highly represented in the transcriptome of N . americanus ( 6 . 1% ) as well as that of A . caninum ( 4 . 6% ) ( see Table 2 ) . These proteases included cysteine , aspartic and metallo- proteases , which are known to function in multi-enzyme cascades to digest haemoglobin and other serum proteins [65] , [66] . In N . americanus , cysteine proteases with high sequence homology to the protein cathepsin B were localized to the gut of adult worms and the corresponding mRNAs shown to be upregulated in the adult stage compared with the infective L3 stage , thus strongly suggesting that these enzymes are involved in blood-feeding [67] . In A . caninum , a cysteine protease ( Ac-CP-1 ) with 86% amino acid sequence identity to those characterized in N . americanus , was shown to be expressed in the cephalic and excretory glands [68] and was detected in the excretory/secretory products ( ES ) [69] of adult worms; thus , it has been proposed that Ac-CP-1 functions as an extracorporeal digestive enzyme at the site of attachment [67] . Another cysteine protease ( Ac-CP-2 ) was localized to the brush border membrane of the intestine and demonstrated to be involved in the digestion of haemoglobin [65] . The N . americanus homologue of Ac-CP-2 ( i . e . Na-CP-2 ) digests haemoglobin [66] and , expressed as a recombinant protein in Escherichia coli and injected subcutaneously into experimental hamsters , has been shown to induce a significant reduction in adult worm burden following challenge infection with L3s of N . americanus [28] , suggesting that the immunogenic response directed against this protein severely impairs the digestion of host proteins by the adult worms . However , recently , a cathepsin-like cysteine protease has been isolated and characterized in the human filarial nematode Brugia malayi and shown by double-stranded RNAi to play an essential role in the early development and maturation of embryos of this nematode [70] . Therefore , it is possible that the abundant transcripts encoding proteases in both adult N . americanus and A . caninum also reflect a key role of these enzymes in embryogenesis . Proteases have also been isolated from larval stages of both A . caninum and N . americanus [71] , [72] . For instance , a metalloprotease in ES of the activated third-stage larvae ( L3 ) of A . caninum has been characterized and demonstrated to be released specifically in response to stimuli that induce feeding [73] . The corresponding cDNA , isolated from an L3 expression library , encoded a zinc-metalloprotease ( Ac-MTP-1 ) of the astacin family , that has been proposed to ( i ) regulate developmental changes associated with the transition from the free-living to the parasitic L3 and the subsequent moult to the fourth-stage larva ( L4 ) [72]; ( ii ) activate host TGF-ß during the infection , which , in turn , could stimulate parasite development directly , determine tissue predilection sites [74] and/or inhibit neutrophil infiltration at the site of penetration [75]; and , ( iii ) facilitate skin penetration or tissue migration by the invading L3 [72] , [76] and/or degrade the cuticular proteins of the sheath surrounding the infective , free-living L3 [77] . In N . americanus , serine proteases have been isolated from ES of the L3 stage and proposed to play a central role in the evasion of the host immune response [71] . Interestingly , a significant number ( n = 135 , 30% ) of N . americanus proteases and protease inhibitors of N . americanus were not predicted to possess signal peptides indicative of secretion ( cf . Tables 1 and 2 ) . The likely explanation for this result is technical and would appear to relate to a 3′-bias in sequence reads [78] , thus affecting the prediction of ORFs as well as the identification of signal peptide sequences at the 5′-ends . Other groups of molecules , such as Ancylostoma-secreted proteins ( ASPs ) , have been proposed to have an immunomodulatory function during the invasion of the host , the migration through tissues , attachment to the intestinal wall and blood-feeding [79] . In the present study , ASPs were amongst the ten most abundant groups of molecules in the N . americanus dataset , and are most abundant in A . caninum ( cf . Table 2 ) . ASPs belong to a large group of proteins , the ‘sperm-coating protein ( SCP ) -like extracellular proteins’ , also called SCP/Tpx-1/Ag5/PR-1/Sc7 ( SCP/TAPS; Pfam accession number no . PF00188 ) , characterized by the presence of a single or double ‘SCP-like extracellular domain’ ( InterPro: IPR014044 ) . In A . caninum , double and a single SCP-domain ASPs , called Ac-ASP-1 and Ac-ASP-2 , respectively , were identified as major components of ES from serum-activated , infective L3s and proposed to be secreted in response to one or more host-specific signals during the infection process [80] , [81] , as also hypothesized in a transcriptomic analysis of serum-activated L3s [19] . In N . americanus , homologues of Ac-ASP-1 and Ac-ASP-2 ( i . e Na-ASP-1 and Na-ASP-2 , respectively ) have been identified in the L3 stage [82]–[84] . Results from crystallography [85] , combined with the observation that Na-ASP-2 induces neutrophil and monocyte migration [86] , suggest that this molecule has a role as an antagonistic ligand of complement receptor 3 ( CR3 ) and alters the immune cascade by preventing the binding of chemotaxin [85] . Because of its immunogenic properties , Na-ASP-2 is under investigation as a vaccine candidate against necatoriasis [7] , [28] , [81] , [87] . In adult A . caninum , at least four other ASPs have been identified to date and named Ac-ASP-3 , Ac-ASP-4 , Ac-ASP-5 and Ac-ASP-6 [72] . Another SCP/TAPS molecule , designated neutrophil inhibitor factor ( NIF ) , has been isolated and shown to play an immunomodulatory role by blocking the adhesion of activated neutrophils to vascular endothelial cells and the subsequent release of H2O2 from activated neutrophils [88] and by interfering with the function of integrin receptors located on the cell surface , which results in the inhibition of platelet aggregation and adhesion [89] . Subsequently , NIF was shown to be transcribed abundantly in the intestines of both A . caninum and N . americanus [34] . The present study revealed that , although highly represented in the transcriptome of adult N . americanus , ASPs were much more abundant in A . caninum ( cf . Results section ) . One of the possible explanations for this finding is that , although the A . caninum dataset was generated from adult worms recovered from their natural host ( i . e . dog ) , the specimens of N . americanus were recovered from a Chinese strain of the golden hamster ( M . auratus ) , which is not a natural host for this parasite [90] , [91] . Indeed , adults of N . americanus recovered from hamsters with patent infections are smaller and less fecund than from the human host [91] . These phenetic differences in this parasite might be associated with variation in transcriptional profiles . However , the difference in prevalence of particular transcripts , such as those of asps , between A . caninum and N . americanus might reflect their distinct roles in the modulation of the host immune response between the two hookworms , an hypothesis that requires testing . A benefit of investigating the transcriptome of parasitic nematodes using predictive algorithms is that potential drug targets can be inferred and/or prioritized . The present study identified a subset of 278 ‘druggable’ proteins , of which 18 did not match any human homologues ( cf . Results section ) . Of these 18 molecules , mitochondrial-associated proteins were significantly represented ( i . e . encoded by the C . elegans orthologues W01a8 . 4 , ucr-1 , F26E4 . 6 and Y71H2aM . 4; cf . Table 4 ) . Mitochondria are essential organelles with central roles in diverse cellular processes , such as apoptosis , energy production via oxidative phosphorylation , ion homeostasis , and the synthesis of haeme , lipid , amino acids , and iron-sulfur ions [92] . In C . elegans , defects in the mitochondrial respiratory chain lead to or are associated with a wide variety of abnormalities , including embryonic , larval and adult lethality , sterility and embryonic defects [92] . Despite their essential roles in numerous fundamental biological processes , knowledge of mitochondrial genes and proteins in parasitic nematodes has been utilized mainly to study their systematics , population genetics and ecology [93]–[95] . However , that some mitochondrial-associated proteins are predicted to be essential in N . americanus and significantly different from human homologues provides a context for the discovery of new drug targets in mitochondrial pathways and chemical compounds that disrupt these pathways [95] , [96] . Amongst the other N . americanus orthologues of essential C . elegans genes , nrs-2 encodes an asparaginyl-tRNA synthetase ( AsnRS ) , which is a class II aminoacyl-tRNA synthetase that catalyzes the attachment of asparagine to its cognate tRNA and is required for protein biosynthesis [97]; loss of nrs-2 function via RNAi has been shown to result in a number of phenotypes , including adult and larval lethality and/or larval arrest [97] . In parasitic nematodes , information on amino acid biosynthesis is limited [98] . Although a number of parasitic helminths , including the nematode Heligmosomoides polygyrus [sic . H . bakeri] and the trematode Fasciola hepatica , have been reported to excrete asparagine during in vitro incubation [99] , [100] , the role of asparagine synthetases in essential biological processes is currently unknown . However , in a study investigating the molecular mechanisms of induced cell differentiation in human pro-myelocytic leukemia , asparagine synthetase transcription was reported to be significantly reduced in maturing monocytes/macrophages [101]; therefore , an active role of asparagine synthetases in the development and growth of cancer cells has been suggested , which led to the hypothesis that the induction of a down-regulation of asparagine synthetases might be a new strategy for the treatment of blast cell leukaemia [102] . This finding raises questions about the role ( s ) of asparagine synthetases in cell differentiation and maturation in parasitic nematodes and the potential of inhibitors of these enzymes as anti-hookworm drugs . The present study has provided new insights into the transcriptome of N . americanus , elucidated similarities and differences between the transcriptomes of N . americanus and the related canine hookworm , A . caninum , and predicted a panel of novel drug targets and nematocides . All except one of the essential ‘druggable’ proteins ( n = 18 ) inferred for N . americanus were present in the A . caninum ( and C . elegans ) but not in the mammalian hosts , suggesting relative sequence conservation for these targets among nematodes . The prediction of such targets is particularly important , considering the risk of emerging drug resistance in parasitic nematodes [102] , [103] . Clearly , transcriptomic and genomic studies , such as that carried out here can facilitate and expedite the prevalidation of targets for nematocidal drugs , although the lack of genomic and transcriptomic data for many nematodes , including the human hookworm A . duodenale , impairs the comparative exploration of essential biological pathways in parasitic nematodes of major public health significance [6] . Furthermore , the present analysis has inferred qualitative and quantitative differences in the transcriptome between N . americanus and A . caninum , raising questions as to the suitability of the latter species as a model for the former . Although these differences require experimental validation , there is a need to define the transcriptome of A . duodenale as a foundation for comparative investigations with a perspective on the identification of new and hookworm-specific drug targets . | The blood-feeding hookworm Necator americanus infects hundreds of millions of people . To elucidate fundamental molecular biological aspects of this hookworm , the transcriptome of adult Necator americanus was studied using next-generation sequencing and in silico analyses . Contigs ( n = 19 , 997 ) were assembled from the sequence data; 6 , 771 of them had known orthologues in the free-living nematode Caenorhabditis elegans , and most encoded proteins with WD40 repeats ( 10 . 6% ) , proteinase inhibitors ( 7 . 8% ) or calcium-binding EF-hand proteins ( 6 . 7% ) . Bioinformatic analyses inferred that C . elegans homologues are involved mainly in biological pathways linked to ribosome biogenesis ( 70% ) , oxidative phosphorylation ( 63% ) and/or proteases ( 60% ) . Comparative analyses of the transcriptomes of N . americanus and the canine hookworm , Ancylostoma caninum , revealed qualitative and quantitative differences . Essential molecules were predicted using a combination of orthology mapping and functional data available for C . elegans . Further analyses allowed the prioritization of 18 predicted drug targets which did not have human homologues . These candidate targets were inferred to be linked to mitochondrial metabolism or amino acid synthesis . This investigation provides detailed insights into the transcriptome of the adult stage of N . americanus . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"genetics",
"and",
"genomics/bioinformatics",
"computational",
"biology/genomics",
"computational",
"biology/transcriptional",
"regulation"
] | 2010 | Massively Parallel Sequencing and Analysis of the Necator americanus Transcriptome |
Different synonymous codons are favored by natural selection for translation efficiency and accuracy in different organisms . The rules governing the identities of favored codons in different organisms remain obscure . In fact , it is not known whether such rules exist or whether favored codons are chosen randomly in evolution in a process akin to a series of frozen accidents . Here , we study this question by identifying for the first time the favored codons in 675 bacteria , 52 archea , and 10 fungi . We use a number of tests to show that the identified codons are indeed likely to be favored and find that across all studied organisms the identity of favored codons tracks the GC content of the genomes . Once the effect of the genomic GC content on selectively favored codon choice is taken into account , additional universal amino acid specific rules governing the identity of favored codons become apparent . Our results provide for the first time a clear set of rules governing the evolution of selectively favored codon usage . Based on these results , we describe a putative scenario for how evolutionary shifts in the identity of selectively favored codons can occur without even temporary weakening of natural selection for codon bias .
The genetic code is redundant with most amino acids encoded by several synonymous codons . In many genomes , some codons are favored over others by selection likely because they are translated more efficiently and accurately [1]–[5] . The selectively favored codons tend to correspond to the most highly expressed tRNAs [6]–[9] . Selection for the use of favored codons should be stronger for genes that are more highly expressed . For this reason , highly expressed genes such as ribosomal genes or translation elongation factors use favored codons almost exclusively and exhibit very high levels of codon bias [6] , [10]–[13] . In contrast , the identity of the codons used by many genes that are not highly expressed may be determined to a large extent by the nucleotide substitution patterns of the genome that are unrelated to natural selection at the level of translation . Previous studies have demonstrated that the overall codon usage patterns of genomes can be predicted based solely on the nucleotide composition of their intergenic regions [14] , [15] . Such studies were interpreted as showing that for most genes selection at the level of translation is only secondary in determining codon usage , as it is too weak to counteract the effects of biases in the patterns of nucleotide substitution that are experienced by the genome in general [14] , [15] . The identity of selectively favored codons varies among organisms [16]–[18] . For example , the favored codon for leucine in Escherichia coli and Drosophila melanogaster is CTG , in Bacillus subtilis TTA , in Saccharomyces cerevisiae TTG , and in Saccharomyces pombe CTT [18] . The rules governing the identities of favored codons in different organisms remain entirely obscure . One possibility is that the optimal codons are chosen randomly in evolution in a process akin to the frozen accident hypothesized to have occurred in the evolution of the genetic code [19] . However , there are some serious difficulties with this possibility . First , some optimal codon choices appear highly structured and counterintuitive . For instance , in Drosophila all optimal codons are G or C ending ( majority are C ending ) while the genome is ∼65% AT rich on average [17] . Even more problematic is the observation that the identity of optimal codons shifts in evolution quite readily . This implies that the frozen accidents of optimal codon choice can become “unfrozen” at times and then after a period of time become frozen again but in a new state . Such shifts would seem to require long periods of weak selection given that they would require a large number of genes to change at a large number of sites seemingly against the pressure of natural selection [1] . One difficulty in gaining insight into this problem is that only few metazoans have clear selection-driven codon bias and the identity of favored codons in other organisms such as bacteria , archea and fungi have not yet been determined . Here we identify the favored codons in 675 fully sequenced bacterial genomes , 52 archeal genomes and 10 fungal genomes ( Text S1 , S2 , S3 ) . We demonstrate that , unlike in Drosophila , the identities of favored codons in bacteria , archea , and fungi correspond to the nucleotide content of the intergenic regions of each genome . Thus , GC rich organisms tend to have GC rich favored codons while AT rich organisms tend to have AT rich favored codons . This indicates that , unlike previously suggested , selection is not secondary in determining the codon usage patterns of genomes . Rather , selection consistently acts in the same direction as the nucleotide substitution biases that determine the nucleotide content of genomes in general . We further use the data in bacteria to demonstrate that once nucleotide substitution patterns are taken into account additional amino-acid specific rules determining the identity of favored codons become apparent . Finally , our findings allow us to suggest a possible mechanism by which the identity of favored codons can change between genomes without necessitating prolonged periods of weak selection on the efficiency and accuracy of translation .
We begin by considering bacterial genomes . A straightforward and widely used way to identify the favored codons is to ask which of the codons encoding a particular amino acid increase in frequency as genes become more biased in the choice of codons overall [13] , [17] , [20] , [21] . Following this reasoning , for each of the 675 bacteria , we calculated the overall degree of codon bias for each gene using the effective number of codons ( Nc ) ( [22] , Materials and Methods ) . Nc measures codon bias of a gene across all codon families without making any assumptions regarding the identity of optimal codons . Values of Nc range between 20 , for extremely biased genes that use only one codon per amino acid , to 61 , for genes that use all synonymous codons equally . A version of Nc , Nc' was suggested by Novembre [23] . Nc' takes into account and adjusts for background nucleotide composition . The intent of Nc' is to define codons that are used unusually frequently given the background GC content of the considered protein coding sequence [23] . For each of the 18 amino acids that are encoded by more than a single codon , we examined the correlation between the frequency of each of its synonymous codons in a gene and the Nc' of the gene . For each amino acid we identified the most favored ( optimal ) codon defined as the codon that showed both the strongest and statistically significant positive Spearman correlation with the overall level of codon bias ( P≤0 . 05/n , where n is the number of codons encoding the amino acid in question , Materials and Methods ) . For some amino acids in some organisms we could find no favored codons . The identities of the identified optimal codons , for each of the 18 amino acids , in each of the 675 bacteria are summarized in Table S1 . Codon bias can be the result not only of selection but also of variation in the patterns of nucleotide substitution . Thus , in order to demonstrate that the codons identified by our procedure are in fact selectively favored , it is necessary to show that variation in codon bias among genes within most genomes cannot be explained without the involvement of selection . To do so , we conducted two tests . First , we examined whether the most codon biased ( MCB ) genes are the most highly expressed genes . Specifically , we asked whether ribosomal genes and translation elongation factors , which are often among the highest expressed genes [24] , [25] , are statistically significantly ( P<0 . 05 ) over-represented among the 100 MCB genes in each genome ( Materials and Methods ) . We found that for 658 of the 675 bacterial genomes studied here this is indeed the case ( Table S2 ) . For most of the bacteria in the study the P-value was much lower than 0 . 05 ( Table S2 ) . This test might be weakened by imperfect annotations in some genomes . Nevertheless , it does show that for the vast majority of bacteria the MCB genes are likely under the strongest selection for optimal codon usage . In order to further demonstrate that codon bias in these genomes is not entirely due to variability in patterns of nucleotide substitutions unrelated to translational selection , we extracted in each genome the first 100 fourfold and twofold degenerate codons of each coding sequence . We then replaced the third codon positions of these coding segments ( CS ) with 100 randomly selected nucleotides from the intergenic sequences adjacent to them , while maintaining to identity of the encoded amino acids . This resulted in a set of intergenic control coding segment ( ICCS ) that maintain the protein sequences and nucleotide content patterns of the genome but remove the effects of selection on synonymous sites that we expect to see in the CS . We calculated the level of codon bias of each of the ICCS and each of the CS and examined for each genome whether the 100 most codon biased CS are significantly more biased than the 100 most codon biased ICCS ( P≤0 . 05 , using a one-tailed Wilcoxon test ) . We found that this is indeed the case for all but one of the 675 bacteria examined . As in the previous test P-values were always much smaller than 0 . 05 ( Table S2 ) . This further suggests that for the vast majority of organisms the optimal codons we identified are indeed likely to be selectively favored . An examination of the identified optimal codons ( Table S1 ) led us to realize that there appears to be a relationship between the identity of optimal codons and intergenic GC content . To examine this relationship systematically we classified the codons in each codon family into the most GC rich , the most AT rich , and those with intermediate GC content ( such codons exist only for Leucine and Arginine ) . We gave a score of 1 to each GC rich codon , a score of −1 to each AT rich codon and a score of 0 to the intermediate codons ( Table S3 ) . For each genome we summed the scores of its optimal codons and divided the sum by the number of codon-families for which we could identify the optimal codon . Thus an organism that has only GC-rich optimal codons will receive a score of 1 while an organism that uses only AT-rich optimal codons will receive a −1 . We plotted these scores against the intergenic GC contents of the genomes ( Figure 1A ) and found a clear correlation between the optimal codon GC score and intergenic GC content ( rspearman = 0 . 88 , n = 675 , P≪0 . 00001 ) . In order to eliminate the possible effects of close taxonomic relationships between some of the analyzed bacteria , we repeated this analysis after randomly selecting a single representative from each bacterial genus . The correlation between the optimal codon GC score and intergenic GC content ( Figure 1B ) remains highly significant ( rspearman = 0 . 84 , n = 263 , P≪0 . 00001 ) . We repeated this analysis for the 52 archea ( Figure 2A and Table S4 ) and the 10 fungi ( Figure 2B and Table S5 ) . We found that for both of these groups there are similar correlations between the intergenic GC content and the optimal codon GC score ( rspearman = 0 . 73 , n = 52 , P<0 . 00001 for archea , rspearman = 0 . 74 , n = 10 , P≤0 . 02023 for fungi ) . Vicario et al . [17] found that D . melanogaster has only GC-rich optimal codons even though the nucleotide substitution patterns of its genome tend towards AT . When we plot the optimal codon score for D . melanogaster ( calculated based on the optimal codons identified in Vicaro et al . [17] ) against the background GC content of D . melanogaster ( estimated in the same paper , based on the sequences of short introns [17] , Figure 2B ) , we find that for its low GC content Drosophila appears to be using a higher proportion of GC rich codons than any of the other three groups of organisms . We also analyzed an additional metazoan , Caenorhabditis elegans , that has a lower optimal codon GC score and a lower GC content [26] than D . melanogaster ( Figure 2B ) . However , there are not enough fully sequenced metazoan genomes with documented selection-driven codon bias to examine the relationship between optimal codon identity and nucleotide content in Metazoa . It is important to note that there is no a priori reason why translationally favored codons should match the nucleotide content of intergenic DNA . Previous studies have demonstrated a relationship between overall codon usage of genomes and their intergenic GC content [14] , [15] . Because in these studies little attention was given to the inner-genome variation in the patterns of codon usage , these results were thought to indicate that selection makes only a weak contribution to creating codon biases , and that the major contributor to the codon bias phenomenon are genome-wide nucleotide substitution biases . By identifying optimal codons and showing that their identity also tracks nucleotide content of intergenic regions we demonstrate that it is not that selection weakly affects codon bias , but rather that it appears to be consistently acting in the same direction as the nucleotide substitution biases of genomes . In order to identify optimal codons , we used Nc' , a measure of codon bias that corrects for variation in genomic GC content [23] . Given our findings it is possible that by using this method we eliminated some of the signal we'd expect to find . For example , based on our findings we expect that the optimal codons in a GC rich genome should be GC rich . Highly expressed genes will use optimal codons more and will be more codon biased and more GC rich . Nc' is expected to correct some of this effect out even though it is in fact true signal rather than noise . Indeed when we identify optimal codons in bacteria using Nc , rather than Nc' we find an even stronger correlation between the GC richness of optimal codons and the GC richness of intergenic sequences ( Figure S1 , rspearman = 0 . 91 , n = 675 , P≪0 . 00001 ) . Interestingly we find that the same optimal codons are almost always identified using both Nc and Nc' for genomes with intergenic GC contents higher than 40% ( Figure 3 ) . However , for genomes with intergenic GC contents lower than 40% the same optimal codon is identified in only ∼50% of cases . In addition we found that our ability to identify optimal codons is much reduced in AT rich genomes . These two findings make sense if selection to use optimal codons is generally weaker for AT rich genomes than for GC rich genomes . Indeed , many of the AT rich bacteria are endosymbionts that are known to be slow growing and in which selection for translation accuracy and efficiency is thought to be weaker [27] , [28] . Even if genomes with GC contents below 40% , for which our ability to clearly identify optimal codons appears to be somewhat reduced are removed from consideration , the correlation between intergenic GC content and the optimal codon GC score remains highly significant ( rspearman = 0 . 73 , n = 366 , P≪0 . 00001 ) . It thus appears that our finding of a relationship between intergenic GC content and the identity of optimal codons is robust to the possible misidentification of optimal codons in the AT rich genomes . To learn more about the rules governing the identity of optimal codons we split all genomes into five groups based on their intergenic GC content . We summarized the identities of the optimal codons in each group for the fourfold degenerate codon families , the codon families with three or six codons , and the twofold degenerate codon families in Figure 4 , Figure 5 , and Figure 6 respectively . To be more certain of our assignment of optimal codons , we demanded that the same optimal codon be identified using both correlations with Nc' and correlations with Nc . If for a certain codon family in a certain genome one or both of these correlations resulted in the identification of no optimal codon , or if they both identified different optimal codons we classify the optimal codon as “none” . Examining these figures allowed us to observe again that GC rich bacteria tend to use GC rich optimal codons while AT rich bacteria tend to use AT rich optimal codons . However , these figures also demonstrate additional rules governing the identity of optimal codons in bacteria . For example , among the fourfold degenerate codons ( Figure 4 ) , for high GC organisms , C is strongly preferred over G in the optimal codons of Threonine , and Glycine . At the same time G appears to be preferred over C in the optimal codons of Proline , and Valine . Our results are less clear for AT rich genomes , as in such genomes for more codon families in more organisms we could identify no clear optimal codon . However , in such genomes , T appears to be preferred over A in the optimal codons of all fourfold degenerate codon families other than Proline . Similarly interesting patterns can be seen for codon families with six members ( Figure 5 ) . For Leucine , for example , in AT rich genomes the TTA codon is preferred among optimal codons . This makes sense as this is the most AT rich codon encoding Leucine . At the same time , for the optimal codons of GC rich bacteria the CTG codon is strongly preferred over the equally GC rich CTC codon . A similar pattern appears for Arginine . For AT rich genomes the optimal codon is most frequently the most AT rich codon ( AGA ) . However , for GC rich genomes CGC is almost always selected over CGG . Such family specific patterns are intriguing and require further study . In a previous study [28] Rocha investigated codon bias from the tRNA perspective by analyzing the copy numbers of the tRNAs with different anticodons in different genomes . Surprisingly , he found that the most frequent anticodons remain constant across different genomes and do not change with GC content . Rocha observed that generally in the first anticodon position ( which will bind to the third codon position ) of twofold-degenerate amino acids , G is always more frequent than A while T is more frequent than C . He therefore expected to observe a preference for C or A in third codon positions of these codon families over G and T [28] . We observe that similarly to other codon families the tendency of organisms to use the more AT rich or GC rich optimal codon out of the two possible twofold degenerate codons depends on intergenic GC content ( Figure 6 ) . However , for codon families that can end in either G or A ( Gln , Glu and Lys ) the shift from using the more AT rich optimal codons to using the more GC rich optimal codons tends to occur at higher GC contents , compared to the codon families that end in either C or T ( Asn , Asp , Cys , His , Phe and Tyr ) . This means that more organisms use the C or A ending codons as expected from Rocha's results . For many organisms only a single tRNA exists for a certain codon family . It is therefore clear that tRNA modifications and wobble rules are involved in allowing a single tRNA to bind different codons . These wobble rules and modifications may be different in different organisms . Such differences made it difficult for Rocha to define expectations as to which codons would be best recognized by the most frequent anticodons in each organism for codon families with more than two members [28] . We could therefore not compare the results of Rocha to our results for such codon families . Our results not only provide a clear set of rules governing the identity of the favored codons , they also provide a possible mechanism by which this identity can shift between organisms . Variation in GC content across genomes implies shifts in nucleotide content . The pattern we found implies that such shifts in nucleotide content are accompanied by shifts in the identity of favored codons . Let us consider a scenario in which a genome begins shifting towards a different global GC content that does not match the GC content of its favored codons . After a while , genes that are not under strong selection at the level of translation will start using codons that correspond to the new GC content of the genome . While , individually these genes may not be expressed highly enough to be under strong selection for the use of favored codons , together they may affect the efficiency of translation substantially . For this reason it may become advantageous for the tRNAs that correspond to these newly frequent codons to increase their expression . While Rocha has shown that the identity of the tRNA with the highest copy number does not tend to change much between bacteria [28] , this can be achieved by increasing the transcription of a certain anticodon tRNA , or through regulation of tRNA modifications . Following this increase , the highly expressed genes will be free to start using the codons that correspond more to the GC content of the genome . This will be encouraged by the new pattern of nucleotide substitutions of the genome and should eventually remove the selection for the high expression of the tRNAs that recognize the old favored codons . As a result after a time new favored codons may emerge that correspond to the nucleotide content of the genome . In order to prove such a scenario it will be necessary to carefully examine shifts in nucleotide content and in the identity of optimal codons across a bacterial phylogenetic tree . In such a way it may be possible to ask whether changes in the identity of optimal codons indeed follow changes in nucleotide content . This analysis is beyond the scope of this paper and so it is important to note that the scenario we suggest here for shifts in optimal codon usage is hypothetical . This scenario is intriguing however as , if true , it explains how the identity of favored codons can shift without requiring a prolonged period of weakened selection . Furthermore , this scenario suggests that while selection for the use optimal codons is strongest for a specific set of highly expressed genes , the identity of the optimal codons is in fact determined largely by the majority of genes , on which selection is much weaker . The codon bias phenomenon has been studied for decades . Yet , basic questions regarding this phenomenon remain unanswered . Here , we provide an insight into one such basic open question: What determines the identity of the codons favored by selection for translation accuracy and efficiency in different genomes . We show that in all three kingdoms of life the identity of the favored codons matches the nucleotide content of the intergenic regions of each genome . Furthermore , once the relationship between the identity of favored codons and nucleotide content is taken into account additional amino-acid specific rules determining the identity of favored codons come to light . We then use our findings to provide a possible answer to a second open question: how can the identity of favored codons shift in evolution and do such shifts require prolonged periods of weakened selection ? Our findings allow us to suggest a scenario for shifts in the identity of favored codons that does not require a weakening of selection .
The completed genomic sequences and coding sequence annotaions of the 675 bacteria , 52 archea , and 10 fungi were downloaded from the NCBI FTP server . ( ftp://ncbi . nlm . nih . gov ) . For each of the fully sequenced bacteria , archea and fungi used in the study ( Text S1 , S2 , S3 ) we extracted the DNA coding sequences of all the annotated proteins . For each protein in each genome we calculated the effective number of codons ( Nc [22] ) . Nc , measures the overall codon bias of a gene across all codon families [22] . The measure does not make any assumptions regarding the identity of the optimal codons . Values of Nc range between 20 , for extremely biased genes that use only one codon per amino acid , to 61 , for genes that use all synonymous codons equally [22] . Since the estimation of Nc is problematic for short sequences , we removed from consideration coding sequences shorter than 50 codons . In order to further account for sensitivity to sequence length , we used the version of Nc supplied by Novembre as part of his ENCprime package that corrects for sequence length [23] . Nucleotide content is also expected to affect Nc . We therefore also used a version of Nc , Nc' which was developed by Novembre and which corrects for nucleotide content [23] . In order to identify optimal codons for a specific genome we calculated for each codon its frequency within its codon family in all of the annotated coding sequences in each genome . We then calculated the correlation between the frequency of each codon within each gene and the overall codon bias ( once using Nc' and once using Nc [23] ) of that gene . We removed from consideration genes in which the codon family appeared less than 10 times . The optimal codon for each codon family was defined as the codon that showed the strongest and significant negative correlation with the Nc or Nc' of the gene . To be considered significant a correlation had to have a P-value smaller or equal to 0 . 05/n , where n is the number of codons in the codon family . In such a way we correct for the fact that we performed more comparisons for more degenerate codon families . Spearman correlations were performed using the R statistical package . In order to randomly select a single member of each bacterial genus , bacteria sharing a genus name ( i . e . Escherichia , or Mycobacterium ) were grouped and a single member of each group was randomly selected . For each genome , we counted how many of the 100 most biased ( lowest Nc ) genes are annotated as “ribosomal” or “elongation factor” . We then randomly selected 100 of the remaining genes in the genome and counted how many of these random genes are annotated as ribosomal genes or elongation factors . We repeated this randomization 1000 times and calculated the P-value that tells us in how many of these random samples does an annotation of “ribosomal” or “elongation factor” appear as often or more often than for the most biased genes . We say that ribosomal genes and elongation factors are significantly over represented among the 100 most biased genes if this P-value is lower or equal to 0 . 05 . To create the intergenic control coding sequences ( ICCS ) we used the following strategy for each of the 675 genomes . I ) We extracted the first 100 four-fold degenerate and two-fold degenerate codons of each protein coding gene . We removed from consideration genes that had less than 100 two-fold and four-fold degenerate codons . II ) For each protein coding gene we extracted its two adjacent intergenic sequences . We concatenated both adjacent intergenic sequences ( the 5′ and the 3′ intergenic sequences ) and selected a 100 base pair segment of this sequence at random . We shuffled the order of the nucleotides of these intergenic segments randomly . We removed intergenic regions shorter than 50 bases and if for a gene there was not at least 100 bases of adjacent intergenic region , we removed that gene from consideration . III ) We created ICCS using the real coding sequences as a backbone and replacing the third codon positions , based on the shuffled adjacent intergenic sequences , while maintaining the encoded protein sequence . For example if in the real protein at the tenth position we have a Valine encoded by the four-fold degenerate codon GTG and the shuffled segment of the adjacent intergenic sequence has a T in the tenth position , our ICCS will have a GTT in the tenth codon position . In the case of a two-fold degenerate codon such as the Lysine codons AA ( A/G ) , we selected AAG if the corresponding intergenic position contained either a G or a C and AAA if the corresponding intergenic position contains an A or a T . At the end of this process we obtained for each genome two sets of coding segments of a consistent length; the “real” coding sequences ( CS ) and the ICCS . Both of these encode exactly the same proteins . The third codon positions of the ICCS reflect the composition of the real gene's adjacent intergenic regions . | Codon bias is a long recognized and long studied biological phenomenon . Yet several basic questions regarding codon usage remain unresolved . Here , we address one such basic open question: the identity of the codons that are favoured by selection for translation accuracy and efficiency varies greatly and , at first glance , idiosyncratically among genomes . What are the rules governing the identity of favoured codons in the different genomes ? We systematically identified the optimal codons of 675 bacteria , 52 archea , and 10 fungi . Using these data , we show that universally across all bacteria , archea , and fungi the identity of the favoured codons tracks the nucleotide content of the genome as a whole . Once the effect of nucleotide content on selectively favored codon choice is taken into account , additional , until now unknown , universal amino acid specific rules governing the identity of selectively favored codons become apparent . Finally , we use our findings to offer a plausible scenario as to how the identity of optimal codons can shift between genomes by tracking the nucleotide patterns of the genome and without necessitating a reduction in selection . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/genomics",
"genetics",
"and",
"genomics/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"molecular",
"biology/translation",
"mechanisms",
"microbiology/microbial",
"evolution",
"and",
... | 2009 | General Rules for Optimal Codon Choice |
Spontaneous waves in the developing retina are essential in the formation of the retinotopic mapping in the visual system . From experiments in rabbits , it is known that the earliest type of retinal waves ( stage I ) is nucleated spontaneously , propagates at a speed of 451±91 μm/sec and relies on gap junction coupling between ganglion cells . Because gap junctions ( electrical synapses ) have short integration times , it has been argued that they cannot set the low speed of stage I retinal waves . Here , we present a theoretical study of a two-dimensional neural network of the ganglion cell layer with gap junction coupling and intrinsic noise . We demonstrate that this model can explain observed nucleation rates as well as the comparatively slow propagation speed of the waves . From the interaction between two coupled neurons , we estimate the wave speed in the model network . Furthermore , using simulations of small networks of neurons ( N≤260 ) , we estimate the nucleation rate in the form of an Arrhenius escape rate . These results allow for informed simulations of a realistically sized network , yielding values of the gap junction coupling and the intrinsic noise level that are in a physiologically plausible range .
Spontaneous activity spreads through neuronal systems of many different mammal species during development . Crucial roles are attributed to this spontaneous activity [1] . Among the most prominent roles is the synaptic refinement in the retina , where spatio-temporally correlated bursts of activity are observed , and it was found that blocking these waves disrupts eye-specific segregation into the visual thalamus [2 , 3] . Therefore , much effort has been devoted in recent years ( e . g . [4–7] ) to understand the mechanisms responsible of retinal waves . The observed patterns of spontaneous activity in the developing retina are remarkably similar across many species [1] . These patterns have been characterized as spatially correlated bursts of activity in the ganglion cell ( GC ) layer , which are followed by periods of silence [8–10] . So far , three different stages of retinal waves have been described in rodents , ( for review see e . g . [1] ) . These different stages are characterized by their underlying circuits , which mature subsequently in development . In stage I , bursts of activity spread between retinal ganglion cells . In this stage , few synapses are identifiable and waves are mediated by gap junctions ( GJs ) and adenosine [11] . Stage II begins with the onset of synaptogenesis and ends with the maturation of glutamatergic circuits while stage III waves end with eyeopening and the onset of vision [12 , 13] . Here , we exclusively focus on the earliest developmental stage ( stage I ) . This stage is prior to the emergence of functional chemical synapses in the retina . Waves show random initiation sites , no directional bias , and a propagation speed of about 450 μm/s . Via patch-clamp recordings , stage I retinal waves were found to be initiated and propagated in the GC layer [11] . In this work we develop a theoretical model of the retina and limit ourselves to a GC layer of bursting neurons which are coupled by GJs . These electrical synapses are formed between each of the major neuron types in the vertebrate retina [14–18] and play a major role in signal processing and transmission of visual information ( for a review , see [18] ) . GJs are formed by two apposed hemichannels , each one formed by an hexameric array of proteins know as connexins . In mammals , connexin-36 and connexin-45 were clearly identified in neurons located in the inner retina [15 , 19] . Both types of connexins follow a distinct expression pattern during retinal development [20] . GJ coupling between neurons has been addressed in various theoretical studies ( see e . g . [21 , 22] ) and has received particular attention in the context of large-scale brain rhythms ( e . g . [23 , 24] ) and traveling wave dynamics ( see e . g . [25 , 26] ) . However , their involvement in the maturation process of the retina is not yet fully understood [27] . GJs have been proposed as the responsible mediator of stage I retinal waves but not yet been used in a model of such waves [5] , which is the problem that we intend to solve with this study . From a physical perspective , GJs act with integration times of the order of milliseconds and were thus argued not to be the mediator of stage I waves [5 , 9] , which are much slower compared to this time-scale . In this work , we present a model of stage I retinal waves , formed by a network of bursting cells . The cells are coupled by the Ohmic currents through GJs which corresponds to the discretized version of a diffusive coupling ( see e . g . [28] for a recent example of complex pattern generation with such a coupling ) ; for recent studies of wave propagation using the alternative spatially extended coupling by an integral kernel , see e . g . [29 , 30] . For our model , we show that under certain conditions , the wave propagation can be sufficiently slow to be the responsible mediator for stage I retinal waves . We discuss analytical estimations of the propagation velocities and compare them to extensive numerical simulations of networks of up to 12 , 000 neurons . Our analytical work , based on diffusively coupled bursting neurons , applies methods from nonlinear dynamics and pattern formation to differential equations with discontinuous resettings . Furthermore , we study the repetitive nucleation of waves caused by noisy input currents and discuss the dependence of the nucleation rates on the noise intensities .
We use the phenomenological Izhikevich neuron model [31 , 32] , known for displaying biologically plausible dynamics . Due to its discontinuous fire and reset mechanism , it is a computationally efficient model of a bursting neuron . Comparable dynamics can be obtained from two-dimensional excitable models such at the Morris-Lecar model , under incorporation of an additional third dimension , e . g . a calcium-dependent potassium current , cf . Sec . 5 . 2 in [33] The model can be regarded as a quadratic integrate-and-fire neuron for the membrane voltage Vi ( t ) of the ith neuron with an additional slow recovery variable ui ( t ) , also referred to as gating variable ( cf . Fig 1 ( a ) for the nullclines of the system ) : τ V d V i d t= a ( V i - V rest ) ( V i - V crit ) - u i + R I i , ( 1 ) τ u d u i d t= b V i - u i , ( 2 ) if : V i ≥ V peak → { V i = V reset , u i = u i + d . ( 3 ) The membrane recovery variable provides negative feedback to the voltage ( cf . Fig 1 ( b ) and 1 ( c ) top ) . The parameters a , b , d as well as Vrest , Vcrit , Vreset , and Vpeak determine the spiking regime of the neuron , with Vrest < Vcrit < Vpeak . The time-scales of the voltage and gating variable are defined by τV and τu , respectively . For u ( t ) ≡ 0 and I ( t ) ≡ 0 , Vrest and Vcrit are the stable and the unstable fixed points of the dynamics , respectively . If Vi ≥ Vpeak , the membrane potential is reset to Vreset , the kth spike time , ti , k , is registered , and the recovery variable is increased by the constant value d . We choose parameters such that the burst characteristics of our model neuron illustrated in Fig 1 roughly agree with experimental measurements from Syed et al . [11] . Specifically , we aim at a burst duration of about 1 − 2 seconds ( cf . Fig 1 ( c ) bottom ) and a spike frequency during bursts of about 5 − 15 Hz . We find those characteristics reasonably met for: a = 0 . 1 , b = 0 . 3 , d = 1 . 2 , τV = 100 msec , τu = 0 . 0003−1 msec , Vrest = −76 mV , Vcrit = −48 mV , Vpeak = 30 mV , Vreset = −50 mV . The bursting mechanism is illustrated in Fig 1 . The chosen time-scale of the gating variable u is comparatively large , but not uncommon for cortical neurons [34] . The total current RIi = R[Igap , i + Inoise , i] is a superposition of the intrinsic noise current and GJ currents from neighboring cells ( see below ) . The intrinsic noise originates from fluctuations of the various channel populations ( sodium , calcium , and different potassium channels , see e . g . [35] ) and is approximated by white Gaussian noise: R I noise , i = τ V 2 D ξ i ( t ) , ( 4 ) with 〈ξi ( t ) 〉 = 0 and 〈ξi ( t ) ξj ( t′ ) 〉 = δij δ ( t − t′ ) and D is the noise intensity . We perform simulations at discrete times with a time step of Δt = 0 . 1 msec according to an Euler-Maruyama integration scheme , see supporting information S1 Text . Ganglion cells are distributed within the ganglion cell layer with a decreasing density towards the outer regions of the retina . For instance , the density in rabbits covers a range from 5000 cells/mm2 down to 200 cells/mm2 ( the mean value is 800 ) [36] . In a previous study of retinal waves observed in rats , Butts et al . [4] used a ganglion cell density of ∼ 4000 cells/mm2 . In their simulations they placed neurons in a regular triangular lattice for which the given density translates to a lattice spacing of 17 μm . Because we focus on the rabbit retina , we assume a triangular lattice with a different lattice spacing of 38 μm , reflecting the lower cell density ( 800 cells/mm2 ) for this system . The reported experimental observations on characteristics of stage I retinal wave were obtained from retina patches of roughly 3 × 5 mm . A mean cell density of 800 cells/mm2 translates to a total cell number estimate of 12 , 000 cells in the studied system . For comparability , we use a similar number of cells for simulations ( i . e . 12 , 100 = 110 × 110 ) . The triangular lattice structure can be seen in Fig 2 ( a ) . Here , we ignore for simplicity the inhomogeneous and irregular structure of the ganglion cell layer . We place N = n × n single ganglion cells in a rectangular domain on a triangular lattice such that every cell is connected with GJs to six nearest neighbors , ( the lattice structure is illustrated in Fig 2 ( a ) ) . For illustrative purposes , we will also consider a one-dimensional chain , in which each neuron has only two neighbors . Because we are interested only in stage I waves , prior to synaptogenesis , these cells are not connected to any other cells , i . e . bipolar and amacrine cells are not part of our model . We choose a common approach ( e . g . [21] ) to model the GJ current as diffusive and instantaneous coupling by R I gap , i = G ∑ n = neighbor ( V n - V i ) , ( 5 ) where G is the rescaled dimensionless GJ coupling , i . e . G = R/Rgap . The membrane resistance R of retinal ganglion cells can experimentally be measured and is in the range of 100-500 MΩ , e . g . [37] . Rgap is the GJ resistance between neighboring ganglion cells in the retina , which depends on the connexin type and the transjunctional voltage difference and is roughly Rgap ≈ 1GΩ [38 , 39] . The values of R and Rgap imply a physiological range for our parameter of G ∈ [0 . 1 , 0 . 5] . Because the time course of the action potential produced by our neuron model is only a coarse approximation of the electrophysiological shape of a spike , the GJ coupling may be stronger or weaker than assumed here . This gives additional justification for choosing a wider range of G . For the two-dimensional setup , we apply two different boundary conditions . For estimating the noise dependence of propagation velocities and nucleation rates , we perform small system simulations ( N∼50-260 ) with periodic boundary conditions in both directions ( system on a torus ) in order to avoid strong finite-size effects . Simulations of the full system with N∼12 , 000 are carried out with two additional layers of neurons on the boundary , that are not exposed to intrinsic noise ( cells on the system boundary have fewer neighbors , between 2 and 5 instead of 6 ) . Neurons in the two outer layers of the large simulations are discarded from all statistical evaluations . Single propagating waves running through the network can be captured by the population activity [40] A ( t ) = 1 N Δ t A ∑ i = 1 N ∑ k ∫ t t + Δ t A d t δ ( t - t i , k ) , ( 6 ) where the index k runs over the spike times of the ith neuron . Hence , A ( t ) is the firing rate , averaged over the network and the time bin ΔtA . We use ΔtA = 0 . 5 seconds , which is comparatively large and covers multiple spikes when the cells are bursting .
If we couple cells in a chain ( as indicated in Fig 2 ( a ) —1D ) and initiate a burst in one of them , we see a propagation of the burst along the chain ( cf . Fig 2 ( b ) ) ; similar voltage traces have also been seen in simulation of computational models of cortex slices , e . g . [41] . A higher propagation speed can be achieved by increasing the GJ conductance parameter G Fig 2 ( c ) . The picture is similar in our two-dimensional setup , for which snapshots are shown in Fig 2 ( d ) . In this case , the wave has been evoked by enforcing a burst in the lower left corner . It propagates as a circularly shaped wave front , which is a consequence of the regularity and rotational symmetry of the system . The gating variable u ( lower row in Fig 2 ( d ) ) can be associated with the experimentally accessible calcium dynamics and resembles calcium fluorescences images [11] . Compared to the membrane potential ( top row ) , the wavefront of the gating variable lags behind , as it slowly builds up during the burst . In both , one-dimensional and two-dimensional simulations in Fig 2 , we have set the intrinsic noise intensity to zero in order to illustrate that wave propagation does not hinge on the presence of fluctuations . We note already here , that the propagation speed in the two-dimensional system matches the order of magnitude of biologically observed values . To determine the speed of the waves from simulation such as shown in Fig 2 ( c ) , we approximate the wave’s shape as circular with a fixed center . We define a wavefront as the group of neurons that spike within the same time bin of Δt = 0 . 1 seconds ( see left illustration in Fig 3 ( a ) ) and measure the front’s mean distance from the center and its mean time instance of occurrence . From the differences of these distances and times , we determine the mean velocity , which we find to be weakly distance dependent , but saturating at about 350 μm from the origin of the wave , cf . Fig 3 ( b ) . In the following , all velocity values are averaged over measurements for the range of distances 350 − 650 μm ( shaded area in Fig 3 ( b ) ) from the point of initiation and we refer to this measuring method as concentric method . The velocities are shown in Fig 3 ( c ) as a function of the GJ parameter for the physiologically relevant range of G ( see Methods ) . We obtain velocities that are in the range of values observed in the rabbit retina [11] , cf . the shaded area in Fig 3 ( b ) . The experimental mean value of about 450 μm/sec is attained for G ≈ 0 . 4 . The propagation and its speed can be theoretically understood as follows . Assuming a steep wave profile , the speed of the wave is given by the inverse of the time it takes a bursting neuron to excite its neighbors , times the displacement of the corresponding wave fronts . We refer to this time as burst onset time difference ( BOTD ) . For simplicity , we neglect noise and consider in the following a one-dimensional setup consisting of three neurons: one initially quiescent neuron ( i ) is connected to a bursting neuron ( i − 1 ) on one side and to a quiescent neuron ( i + 1 ) on the other side . They are separated by the lattice spacing ℓ = 38 μm , hence the velocity is defined as v1D = ℓ/TB . Therein , TB denotes the analytical approximation of the BOTD for this one-dimensional case . The approximation TB for the BOTD between neighboring neurons can be derived using three assumptions ( details in S1 Text ) . First , we assume a constant gating variable ( u ( t ) ≈ ur = const ) , which is reasonable on a short time-scale , because τu ≫ τV . Second , we replace the voltage variable of the bursting neuron Vi−1 ( t ) by its temporal average V ¯ b = const , that can be analytically calculated ( see S1 Text ) and for our standard parameters is V ¯ b = - 34 mV . Third , we replace the voltage of the quiescent neuron that is not directly connected to the bursting neuron by the resting potential , Vi+1 = Vr . Consequently , the GJ current seen by the driven neuron reads R I gap , i = G ( V i - 1 + V i + 1 - 2 V i ) ≈ G ( V ¯ b + V r - 2 V i ( t ) ) , and the resulting dynamics until the voltage Vi reaches the peak potential for the first time is effectively one-dimensional and can be recast to the form ( cf . details in S1 Text ) : τ V d V i d t ≈ a ( V i - V rest ) ( V i - V crit ) - u r + G ( V ¯ b + V r - 2 V i ) . ( 7 ) This first order ordinary differential equation can be solved via separation of variables to find t ( V ) . We obtain it by first calculating the difference of the times from the voltage being at its peak potential and its resting potential . However , the driven neuron is already exposed to the driving GJ current while the voltage of the bursting neuron travels to its first spike time ( cf . Fig . A of S1 Text ) . Therefore , for simplicity we subtract the first inter-spike interval TISI from the beforehand calculated time difference: T B ( G ) =t ( V peak ) - t ( V r ) - T ISI . ( 8 ) The explicit expression is lengthy and derived in S1 Text , resulting in Eq . O of S1 Text . Comparing TB to simulations of a one-dimensional chain shows a reasonable agreement ( cf . Fig . A of S1 Text ) , although the theory overestimates the simulated values , in particular , for larger values of G . For comparison we also discuss a corresponding result for the wave velocity in the continuum limit in S1 Text . In the two-dimensional setup at larger times , the wave attains a planar shape as indicated in Fig 3 ( a ) , where red circles represent bursting neurons and blue and yellow circles represent driven and quiescent neurons . Now , we assume that the wave front is perfectly flat and all neurons shown in the same color share an identical voltage . In that case , the propagation mechanism simplifies to two bursting neurons exciting one quiescent neuron , whose membrane potential is further affected by two quiescent neurons . Hence , we can mimic the quasi one-dimensional situation by doubling the value of G and additionally taking into account the modification of the effective length , i . e . ℓeff = ( 3/4 ) 1/2 ℓ , see Fig 3 ( a ) . Consequently , we can approximate the velocity in the two-dimensional system as v 2 D ( G ) =3 / 4 · ℓ T B ( 2 G ) . ( 9 ) Calculated velocities v2D ( G ) are shown in Fig 3 ( c ) by the blue line , underestimating the true velocity ( circles ) but providing a correct order-of-magnitude estimate . Note that so far we restricted the considerations to a purely deterministic setup . Our simulations with noise indicate that moderate fluctuations have only little impact on the mean velocities . In the stochastic version of our system , we observe spontaneous waves that resemble those found in experiments [11] . Experimentally , it was observed by Syed et al . [11] that the spontaneously nucleated waves appear with a mean inter-wave interval TIWI of 36 seconds . In our model , waves are initiated by noise , since neurons are set in the excitable regime and cannot generate periodic spiking or bursting without external input . We expect that the nucleation rate per neuron depends strongly on the noise intensity D . To characterize this dependence , we simulate small systems ( N∼50-260 , see Methods ) with periodic boundary conditions for two different values of GJ coupling and different noise intensities , cf . Fig 4 . With the understanding that every neuron has the same chance to trigger a wave , the global nucleation rate should be linear with N to a first approximation . Thus we measure the nucleation rate per neuron as r = 1/ ( TIWI N ) . As demonstrated in Fig 4 by the linear dependence of the rate’s logarithm on the inverse noise intensity , we obtain an Arrhenius rate r=r 0 exp ( - Δ U / D ) . ( 10 ) The effective potential barrier ΔU depends on G and the system size N and saturates for sufficiently large systems ( inset ) for both values of G . The increase of the potential barrier with G can be understood to first approximation by the effective change of the current-voltage relation in the single neuron . The GJ coupling term Eq ( 5 ) leads to an effective increase in the leak current that stabilizes the resting potential and makes it harder to initiate a burst . This mechanism is dominant in comparison to the influence of other coupling effects and the stochasticity of the neighbors on the nucleation rate ( supported by additional simulations , see Fig . B of S1 Text ) . The more subtle dependence of ΔU on the system size can be explained as follows: Coupling stochastic neurons in small systems with periodic boundary conditions leads to spatial correlations and thus effectively to stronger noise . This effect can be neglected for large system sizes or weak coupling , but has a measurable effect otherwise ( cf . Fig 4 and Fig 4 inset ) . Our results so far can be used to predict the mean inter-wave interval and the propagation speed of retinal waves for a system size N = 12 , 100 that roughly corresponds to the experimentally studied patch size in Ref . [11] . Vice versa , we can infer an approximate value of the noise intensity D that leads to the experimentally observed value of TIWI = 36 seconds and test this in numerical simulations of the full system . For our estimation of the rough value of the noise intensity in a large system , we have to take into account that the single neuron undergoes a substantial refractory period of Tref ≈ 14 seconds after bursting ( estimated from small-system simulations investigating the minimal mean inter-wave interval for various noise intensities ) . The mean inter-wave interval is then given by TIWI = Tref + 1/[N ⋅ r ( D ) ] and the estimated value of the noise intensity follows from the Arrhenius law , Eq ( 10 ) , as D *=Δ U / ln [ N ( T IWI - T ref ) r 0 ] ≈ 0 . 050 ( 11 ) ( for G = 0 . 4 , and r0 = 6 and ΔU = 0 . 71 , fit parameters from Fig 4 , solid line with N = 256 ) . The estimated parameters , G = 0 . 4 and D = 0 . 050 , can now be used in a large-scale simulation . In Fig 5 ( a ) , we show snapshots of the full system’s gating variable ( a proxy for the experimentally accessible calcium concentration ) . The wave front seen in the experimentally observable area ( box in Fig 5 ( a ) ) looks similar to experimental measurements , cf . Ref . [11] . From Fig 5 ( b ) , it becomes evident that the mean inter-wave interval becomes much shorter for a slight increase in D . The mean inter-wave interval at these parameter values is not exactly 36 seconds , but somewhat larger: these statistics depend very sensitively on the value of the noise intensity ( i . e . on the second leading digit , cf . Fig 5 ( c ) middle ) . This is seen in the global population activity , that reveals a wave going through the system as a single peak vs . time . The dependence of crucial neural statistics on the noise intensity is illustrated in Fig 5 ( c ) . In contrast to the mean inter-wave interval , the mean velocity of the wave does not depend strongly on the noise ( Fig 5 ( c ) , top ) but stays close to the experimentally observed mean value ( dashed line ) . This is due to the fact , that the wave , once it is initiated , is largely determined by the deterministic propagation mechanism explained above . The fine tuning of the noise intensity shows that the experimental value of 〈TIWI , exp〉 = 36 seconds is attained for a noise level of D = 0 . 052 , slightly larger than D* ( estimated above ) . How realistic is this noise level ? To address this question , we show at the bottom of Fig 5 ( c ) the standard deviation of the subthreshold voltage fluctuations , σV , as a function of the noise intensity D . σV increases only slightly with D and attains values around 1 . 6 mV . To our knowledge , there are no detailed investigations of intrinsic noise sources in retinal ganglion cells at embryonic age . Because in this developmental stage there are no chemical synapses present [42] , the synaptic background fluctuations can be excluded for our system: in the recurrent networks of the cortex , fluctuations stem mainly from the many synaptic interactions among the neurons , resulting in the famous asynchronous irregular state [43] that can be highly variable [44–46] . In our system , one likely source of variability is channel noise that typically leads to small membrane potential fluctuations with a standard deviation σV below 0 . 6 mV [47 , 48] . The noise intensity that is required for the experimentally observed inter-wave interval results in sub-threshold voltage fluctuations that are three times bigger , cf . Fig 5 ( c ) bottom , suggesting that besides ion channel noise there are additional sources of fluctuations present . These could result from stochasticity of GJs itself but also indirectly from GJs via differences in individual resting potentials ( for the heterogeneity of the resting potential in similarly sized cells , pyramidal cells in the cortex , see [49] ) . In any case , the apparent voltage fluctuations of about 1 . 6 mV are well within the range of experimentally observed voltage noise in embryonic ganglion cells ( cf . Fig . 1 in Ref . [11] ) . The investigations presented in this paper propose a GJ-based model of stage I waves in the developing retina . Starting with a neuron model that roughly reproduces the spiking properties of a burst of one single retinal ganglion cell , we incorporated GJ coupling of physiologically plausible strength and temporally uncorrelated fluctuations . This allowed us to reproduce the characteristics of wave nucleation and slow wave propagation in the early retina . Earlier it was believed that GJs can play a role in fast neural transmissions only [5 , 9] , since the current in electrical synapses responds much quicker than neurotransmitters in chemical synapses . As shown in our paper , however , it is possible to obtain a limited transmission speed in a simple Ohmic model of the GJ coupling . Furthermore , although stochastic fluctuations are strong enough to ignite bursts with the correct nucleation rate , they do not distort the propagating fronts very much , i . e . the wave propagation is still a reliable process . The reason for the slow transmission we observe can be found in the nonlinear dynamics of the single neuron . The Izhikevich model that we use for the ganglion cell is essentially a quadratic integrate-and-fire neuron model with a slow adaptation variable . This model is the normal form of a saddle-node bifurcation and has a pronounced latency if close to this bifurcation , i . e . the spike response to a current step ( in our case provided by a neighboring bursting cell ) is considerably delayed because the system experiences the “ghost of the former fixed point” , see Ref . [50] . The presence of weak noise modifies this picture only slightly [51] . Although our model accounts for the most important features of wave nucleation and propagation for stage I retinal waves , it cannot explain the strong variability of the experimentally measured statistics ( error of velocity ±91 μm/sec [11] ) . This is due to a number of model simplifications , which we now concludingly discuss . Firstly , the real system is much more heterogeneous than in our model , both with respect to the lattice structure as well as with respect to the local coupling between cells; secondly , GJs may couple more than next neighbors and their conductivity may be noisy and voltage gated; thirdly , the detailed dynamics of ganglion cells is certainly more complex than can be captured by the Izhikevich model; last but not least , the white Gaussian noise in our model is a rather coarse approximation of the channel noise and other fluctuations in the system . In our model , we arranged the neurons on a highly regular lattice with a cellular spacing according to an experimentally determined mean value of cell density , neglecting the strong heterogeneities in the distribution [36] . On this lattice , each cell is connected to exactly six nearest neighbors . Given the aforementioned heterogeneity , the numbers and distances between neighbors will be more broadly distributed than in our model . Incorporating these heterogeneous features in the simulations would likely broaden the range of observed velocities and thus better reflect the considerable variability found in experimentally measured values . The soma size of ( rabbit ) retinal ganglion cells ( < 30μm , e . g . Ref [36] ) is smaller than our employed lattice spacing , implying GJ coupling between dendrites rather than soma-soma coupling only . The size of the dendritic arbor of retinal ganglion cells is ∼ 100 − 130μm , thus suggesting direct communication between cells that are up to the threefold of the lattice spacing apart . In our simulations with only next-neighbor coupling , we could reproduce the experimentally observed velocity with a comparatively large coupling constant of G = 0 . 4 ( physiological range was G ∈ [0 . 1 , 0 . 5] , see Methods ) . It is conceivable , that this large G value is an effective description of a system with larger effective GJ neighborhood but with a smaller ( and possibly distance-dependent ) coupling value G . Put differently , we expect similar results for the wave speed in a system with extended coupling neighborhood but reduced coupling strength per connection ( with the latter still being within the physiological range ) . Regarding the neuron model and the incorporation of noise , we note that for developed retinal ganglion cells detailed multi-compartment conductance-based models with stochastic ion channels exist [35] . With more electrophysiological data available , it will certainly be possible to develop biophysically more realistic models of the bursting ganglion cell at the early stage . Furthermore important for our problem will be the incorporation of stochastic models of GJs [52] with voltage-dependent kinetics [53 , 54] and the heterogeneity of physiological parameters such as the resting potential . Such detailed models are certainly difficult to simulate for large networks but could be employed to estimate the total noise intensity in the system and to identify the dominant noise source , cf . similar approaches in Refs . [35 , 55 , 56] . | Retinal waves are a prominent example of spontaneous activity that is observed in neuronal systems of many different species during development . Spatio-temporally correlated bursts travel across the retina at a few hundred μm/sec to facilitate the maturation of the underlying neuronal circuits . Even at the earliest stage , in which the network merely consists of ganglion cells coupled by electric synapses ( gap junctions ) , it is unclear which mechanisms are responsible for wave nucleation and transmission speed . We propose a model of gap junction coupled noisy neurons , in which waves emerge from rare stochastic fluctuations in single cells and the wave’s transmission speed is set by the latency of the burst onset in response to gap junction currents between neighboring cells . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] | [
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"nervous",
"system",
"membrane",
"potential",
"ocular",
"anatomy",
"condensed",
"matter",
"physics",
"junctional",
"complexes",
"electrophysiology",
"neuroscience",
"gap",
"junctions",
... | 2019 | Gap junctions set the speed and nucleation rate of stage I retinal waves |
The lysin LysGH15 , which is derived from the staphylococcal phage GH15 , demonstrates a wide lytic spectrum and strong lytic activity against methicillin-resistant Staphylococcus aureus ( MRSA ) . Here , we find that the lytic activity of the full-length LysGH15 and its CHAP domain is dependent on calcium ions . To elucidate the molecular mechanism , the structures of three individual domains of LysGH15 were determined . Unexpectedly , the crystal structure of the LysGH15 CHAP domain reveals an “EF-hand-like” calcium-binding site near the Cys-His-Glu-Asn quartet active site groove . To date , the calcium-binding site in the LysGH15 CHAP domain is unique among homologous proteins , and it represents the first reported calcium-binding site in the CHAP family . More importantly , the calcium ion plays an important role as a switch that modulates the CHAP domain between the active and inactive states . Structure-guided mutagenesis of the amidase-2 domain reveals that both the zinc ion and E282 are required in catalysis and enable us to propose a catalytic mechanism . Nuclear magnetic resonance ( NMR ) spectroscopy and titration-guided mutagenesis identify residues ( e . g . , N404 , Y406 , G407 , and T408 ) in the SH3b domain that are involved in the interactions with the substrate . To the best of our knowledge , our results constitute the first structural information on the biochemical features of a staphylococcal phage lysin and represent a pivotal step forward in understanding this type of lysin .
Although Staphylococcus aureus is a common habitant of the human skin and respiratory tract , several highly pathogenic strains are major causes of hospital-associated infections and can be life threatening , particularly in immunocompromised patients [1] . Over the past three decades , the incidence of methicillin-resistant S . aureus ( MRSA ) infection , particularly as caused by community-associated MRSA ( CA-MRSA ) isolates , has dramatically increased worldwide , which raises serious concerns within the medical community [2] . USA300 is the most prevalent CA-MRSA strain and accounts for up to 97% of all CA-MRSA infections [3] . The treatment of infections caused by CA-MRSA has become increasingly difficult due to the emergence of multidrug resistance [4] . Therefore , an urgent need exists for novel therapeutic agents directed against this formidable pathogen [5] , [6] . Lysin ( also known as endolysin ) is a cell wall hydrolase that is synthesized at the end of the phage lytic life cycle and is involved in cell lysis and the release of progeny particles from host cells [7] . Lysin can also rapidly and specifically lyse Gram-positive bacteria when exogenously applied [8] . Because the bacterial cell wall is conserved and is necessary for the life cycle , the current lack of reports on the development of bacterial resistance against lysin is not surprising [9] . Additionally , the species- or type-specificity guarantees that the lysin will not affect the normal microflora [10] . Thus , lysin is thought to be a promising potential antibacterial agent . In our previous study , we reported that LysGH15 , which is encoded by the staphylococcal phage GH15 , demonstrates strong lytic activity against MRSA in vitro and in vivo [11] , [12] . LysGH15 shares very high sequence identity with other lysins of class III staphylococcal phages , such as LysK , phi11 , and MV-L [13] . Moreover , these lysins possess a modular structure containing an N-terminal CHAP domain ( cysteine , histidine-dependent amidohydrolases/peptidases ) , a central amidase-2 domain ( N-acetylmuramoyl-L-alanine amidase ) , and a C-terminal SH3b domain ( the SH3 bacterial binding domain , which typically contains 60–70 residues and is homologous to eukaryotic SH3 proteins ) [14]–[17] . CHAP domains as well as amidase-2 domains are types of catalytic domains commonly found in lysins . Although the potent lytic activity against MRSA has been extensively characterized , the molecular mechanism of LysGH15 and of other homologous staphylococcal phage lysins has remained unclear . In this study , we report that the lytic activity of the LysGH15 CHAP domain is critically dependent on calcium ions . To elucidate the molecular mechanism , we determined the structures of three individual LysGH15 domains . The crystal structure of the CHAP domain unexpectedly reveals a calcium-binding site near the Cys-His-Glu-Asn quartet active site ( the active site that consists of cysteine , histidine , glutamic acid , and asparagine typically found in members of the CHAP family [18] ) , and site-directed mutagenesis further confirms that both the calcium ion binding site and the conserved Cys-His-Glu-Asn quartet are required for the lytic activity of the LysGH15 CHAP domain . Structure-based mutagenesis also confirms that E282 and the zinc ion play an important role in maintaining the lytic activity of the amidase-2 domain . Furthermore , NMR titration-guided mutagenesis identifies the potential target binding interface of the LysGH15 SH3b domain . These details provide a pivotal step forward in understanding this type of lysin .
Bioinformatic analysis ( PSIPRED [19] ) and known domain boundaries led us to define the CHAP construct as residues 35–160 , the amidase-2 construct as residues 197–346 , and the SH3b construct as residues 412–481 in LysGH15 ( Figure 1 ) . According to this information , we designed several constructs of each individual domain . Some constructs could be expressed and purified as soluble proteins in Escherichia coli . Each individual domain eluted as a single peak during size-exclusion chromatography ( SEC ) ( Figure S1A ) . Sedimentation velocity experiments using analytical ultracentrifugation ( AUC ) indicated that the three individual domains all exist as monomers in solution ( Figure S1B ) . The activity of each individual catalytic domain was determined ( Table 1 ) . The wild type ( wt ) CHAP domain alone also demonstrates bactericidal activity , but this activity is much weaker than that of full-length LysGH15 , and a high concentration ( 50 µM ) is required to achieve a similar bactericidal effect as the full-length LysGH15 ( 0 . 25 µM ) . At concentrations below 1 µM , nearly no bactericidal activity for the CHAP domain is detected . Moreover , the lytic activity of the isolated CHAP domain requires a longer time , in contrast to the extremely rapid lysis of the CA-MRSA strain USA300 by full-length LysGH15 . The amidase-2 domain of LysGH15 does not demonstrate lytic activity . Surprisingly , the amidase-2 domain is able to enhance the lytic activity of the CHAP domain against USA300 ( Table 1 ) . PSI-BLAST analysis using the National Center for Biotechnology Information ( NCBI ) database reveals that the amidase-2 domain contains a conserved zinc-binding site . To determine the role of the zinc ion in the activity of the amidase-2 domain , the combination of the CHAP and amidase-2 domains was pretreated with EDTA ( 1 mM ) , and excess EDTA was removed by dialysis . Surprisingly , the lytic activity of the combination was completely abolished . Additionally , the individual CHAP domain is also sensitive to EDTA ( Figure 2 ) , which suggests that the lytic activity of the CHAP domain is dependent on the presence of metal ions . However , the PSI-BLAST analysis does not indicate a conserved ion-binding site in the CHAP domain or in homologous proteins . Therefore , we sought to determine the structure of LysGH15 , particularly that of the CHAP domain , and further elucidate the nature of this phenomenon . Full-length LysGH15 was expressed in E . coli , but the protein exhibited degradation . Although the introduction of a C-terminal His-tag improved the stability of LysGH15 , the slow degradation of LysGH15 was not amenable to crystallization , which hindered our initial attempts to crystallize full-length LysGH15 . However , the crystallization of individual LysGH15 domains was possible , and the respective structures were successfully determined ( Table S1 ) . The X-ray crystal structures of the CHAP and amidase-2 domains were determined using the selenium single-wavelength anomalous dispersion ( Se-SAD ) and iodide single-wavelength anomalous dispersion ( I-SAD ) methods , respectively , due to the lack of homologous structures . Crystals of the individual CHAP domain encompassing residues 1–165 and of the amidase-2 domain encompassing residues 165–403 were obtained , as shown in Figure 1 . Additionally , the three-dimensional structure of the SH3b domain was determined using NMR spectroscopy in solution ( Table S2 ) . The NMR spectra of the SH3b domain were acquired using a construct containing residues 368–495 ( Figure 1 ) . The CHAP domain crystallizes in the P6222 space group with two molecules in the asymmetric unit ( ASU ) . Iterative rounds of model building followed by refinement result in a model with good statistics and geometry ( Table S1 ) . The quality of the electron density permits the unambiguous modeling of residues 1–164 . The final model is refined to 2 . 69 Å resolution ( Rwork = 17 . 37% , Rfree = 20 . 39% ) . The two molecules form a dimer that is mediated by hydrogen bonding interactions at the central interface by one molecule of Bis-Tris-propane that was present in the crystallization reservoir solution . However , a Bis-Tris-propane molecule does not mediate homodimer formation of the CHAP domain in solution and cannot affect the activity of the CHAP domain . The CHAP domain forms a globular structure that is comprised of three α-helices that are packed against six β-sheets , as shown in Figure 3A . Notably , as shown in Figure 3B , residues D45 , D47 , Y49 , H51 , and D56 coordinate a central Ca2+ ion that forms the classical 12-residue ( positions 1 , 3 , 5 , 7 , and 12 ) calcium-binding site . The presence of the calcium was further confirmed using inductively coupled plasma atomic emission spectrometry ( ICP-AES ) analysis . The side chains of D45 , D47 , and D56 and the main chains of Y49 and H51 are 2 . 36 , 2 . 37 , 2 . 13 , 2 . 32 , and 2 . 27 Å , respectively , from the Ca2+ ion , as shown in Figure 3B and Table 2 . Additionally , the coordination sphere of the calcium-binding site is completed by a water molecule ( Figure 3B ) . The isothermal titration calorimetry ( ITC ) demonstrated that the equilibrium dissociation constant of the LysGH15 CHAP domain for Ca2+ is approximately 27 µM ( Figure S2 ) . The circular dichroism ( CD ) spectroscopy showed that the presence/absence of Ca2+ does not affect the secondary structures of the LysGH15 CHAP domain ( Figure S3A ) . However , the fluorescence-based thermal shift assays showed ∼2°C shift in Tm of the CHAP domain with/without Ca2+ , which indicated that Ca2+ has a slight contribution to the protein thermostability ( Figure S3B ) . Sequence-based searches reveal that the CHAP domain shares very little identity ( <28% ) ( Figure 4A ) with the proteins that are deposited in the Protein Data Bank ( PDB ) ; therefore , the CHAP family represents a rare example of a protein family that is defined by a unique family member . In contrast , searches for structurally similar proteins in the PDB using the DALI server [20] produces several hits of sizeable Z-scores , as shown in Table 3 . The most structurally homologous protein is the CHAP domain of PlyC [21] , which is a streptococcal-specific phage lysin ( Z-score = 11 . 4 , Root-mean-square deviation ( RMSD ) = 2 . 32 Å ) . Additionally , the LysGH15 CHAP domain also possesses a fold that is similar to the structures of Staphylococcus saprophyticus SSP0609 and several other proteins ( Figure 5A , Table 3 ) . The superposition of the LysGH15 CHAP domain with these proteins indicates the putative peptidoglycan-binding groove as demonstrating the highest similarity , particularly the Cys-His-Glu-Asn quartet ( Figure 5B ) . Surprisingly , none of these structurally homologous proteins contains a calcium-binding site corresponding to the position of the calcium-binding site in the LysGH15 CHAP domain ( Figure 5B ) , which indicates that the calcium-binding site of the LysGH15 CHAP domain is unique . Therefore , the CHAP domain of LysGH15 represents a sub-family of the CHAP family . The sequence and structure of the calcium-binding site in the LysGH15 CHAP domain were compared with the 12-residue calcium-binding sites from randomly selected proteins in the PDB . As shown in Figure 3C , residues at positions 1 and 3 are highly conserved and are frequently aspartate , and the residue at position 12 is generally glutamate ( an aspartate in the LysGH15 CHAP domain ) , whereas positions 5 and 7 demonstrate greater deviation among these proteins . From the structures shown in Figure 3D , the initial half of the “annulus” loops is similar among the different proteins , whereas the terminal half of the loops exhibit large deviations . To identify the metal ion that critically affects the lytic activity of the CHAP domain , common metal ions were added to the EDTA-inactivated CHAP domain . Interestingly , we found that the loss of the lytic activity can be restored only when calcium is added , whereas the addition of other ions , such as magnesium , iron , or zinc , is not able to activate the lytic activity of this domain ( Figure 2 ) . This finding indicates that the lytic activity of the CHAP domain is specifically dependent on calcium . An identical behavior was also detected for LysGH15 , which led to the question of whether the calcium ion that critically affects the lytic activity is coordinated by these five residues . To answer this question , residues D45 , D47 , Y49 , H51 , and D56 were individually mutated to alanine . The lysis assay indicated that the D45A , D47A , and D56A mutations all result in a significant loss of bactericidal and cell wall catalytic activity , as shown in Figure 3E and Figure 3F . Moreover , a unique calcium spectrometry signal is not detected in the D45A , D47A , or D56A mutant proteins using ICP-AES ( Table 2 ) . Furthermore , supplementation with calcium does not restore the activity of these three mutants . In contrast , the activities of the Y49A and H51A mutants marginally decrease ( i . e . , these mutants retain >80% of the activity ) . Additionally , the Q53A mutation was constructed as a control that retains lytic activity . Collectively , these results indicate that the calcium ion bound by these five residues is necessary for the lytic activity of the CHAP domain , and residues D45 , D47 , and D56 play an important role in coordinating this calcium ion . Additionally , analysis of the molecular conservation of the CHAP domain using ConSurf [22] indicates several highly conserved residues at or near the surface , such as D47 , Q53 , C54 , D56 , G74 , N75 , H117 , E134 , and N136 , which form a narrow and deep groove ( Figure S4A ) . The architecture and size of this groove indicate that it most likely serves as the binding site for a portion of the peptidoglycan . In the center of the groove , C54 , H117 , E134 , and N136 form the Cys-His-Glu-Asn quartet . C54 is positioned at the beginning of helix α1 , followed by H117 at the beginning of strand β3 , E134 at the end of strand β4 , and N136 proximal to the beginning of the loop that links strands β4 and β5 ( Figure 3B ) . To assess the roles of these four residues , mutated LysGH15 CHAP domains were expressed and purified . We found that the C54A , C54S , H117A , and N136A mutations in the CHAP domain result in a loss of bactericidal and cell wall hydrolytic activity . The mutant E134A only retains a portion of the lytic activity ( E134A retains ∼58% activity in the cell wall and ∼15% in cells ) , as shown in Figure 3E and Figure 3F . These results indicate that this quartet plays an important role in the lytic activity of the CHAP domain . Coincidentally , the calcium ion that is present within this active site lies particularly close to C54 . However , ICP-AES analysis indicates that the C54A and C54S mutations do not affect the calcium binding of the CHAP domain . Moreover , a C54S mutation in the full-length LysGH15 results in the complete loss of lytic activity , as shown in Table 1 . Therefore , the CHAP domain primarily determines the lytic activity of LysGH15 . However , the C54S LysGH15 mutant ( 0 . 25 µM ) does not demonstrate lytic activity even with complementation using the native CHAP domain ( 0 . 25 µM ) . It is likely that the high lytic activity of LysGH15 cannot be ascribed to the CHAP domain alone . Additionally , D45A , D47A , or D56A mutations in LysGH15 also result in the complete loss of lytic activity ( Table 1 ) , which further indicates that the lytic activity of LysGH15 is dependent on the calcium . The structure of the amidase-2 domain ( residues 165–403 ) was determined at 2 . 27 Å resolution . As shown in Figure 6A , the amidase-2 domain exhibits a ββααββαβααααα topology . Electron density for 24 residues at the N-terminus and 30 residues at the C-terminus is absent . A recessed area located on the surface of this structure is enclosed by helices α2 , α3 , α5 , α7 and several loops . β2 , β3 and β5 form the bottom of this recessed area . One zinc ion is located at the center of this groove and interacts with the side chains of residues H214 , H324 , and C332 , as shown in Figure 6B . The tetrahedral coordination sphere of zinc is completed by a water molecule , which is an arrangement often observed in the active sites of zinc-dependent metalloenzymes . The architecture and size of the groove indicate that it most likely serves as the binding site for a portion of the peptidoglycan ( Figure S4B ) . This potential active site is solvent exposed and lies within a shallow groove on the protein surface , which is consistent with its ability to cleave a highly crosslinked and branched polymer . We compared the structure of the LysGH15 amidase-2 domain with other proteins available in the PDB , as shown in Figure 7A . Despite similar folds , the amino acid sequences of these proteins share very little identity ( <25% ) with the LysGH15 amidase-2 domain . Indeed , the closest protein is the amidase domain of PlyL from Bacillus phage ( PDB ID: 1YB0 ) [23] , which shares only approximately 20% sequence identity ( Figure 4B ) and an RMSD value of 1 . 62 Å for the LysGH15 amidase-2 domain ( Table 3 ) . Several proteins that exhibit structural similarity with the LysGH15 amidase-2 domain are summarized in Table 3 . All of these structures share the α1–α5 helices and all of the β-strands ( β1–β5 ) . In particular , β2 , β5 , α5 , and Lα5α6 are well conserved because they contain the zinc-binding sites and conserved residues , including N275 and E282 . Mutation of the three zinc-binding residues to alanine results in a complete loss of the activity ( i . e . , the ability to enhance the lytic activity of the CHAP domain ) of the LysGH15 amidase-2 domain , as shown in Figure 6C . E282 in the LysGH15 amidase-2 domain lies in an identical position as E90 in PlyL and E93 in xlyA ( Figure 7B ) , which places it in a subclass of the N-acetylmuramoyl-L-alanine amidases [23] . Site-directed mutagenesis indicates that the activity of the E282A mutant is completely lost ( Figure 6C ) . These results indicate that both the zinc-binding site and E282 are necessary for the activity of the LysGH15 amidase-2 domain . However , the E282A mutation or mutation of the zinc-binding residues ( H214A , H324A , or C332A ) does not affect the lytic activity of full-length LysGH15 ( Table 1 ) . T330 in LysGH15 is structurally homologous to K135 in PlyL and K128 in the T7 lysozyme ( histidine in AmiE ) . The finding that the T330A mutant demonstrates a 50% decrease in enhancing ability for the CHAP domain suggests that T330 is also involved in the activity of the amidase-2 domain . N275 in the LysGH15 amidase-2 domain is an additional conserved residue , which corresponds to N112 in the autolysin AmiE . Although the construct of the SH3b domain contains residues 368–495 , the determined NMR structure indicates that residues 400–495 form a compact domain , whereas residues 368–399 form a flexible linker . The SH3b domain consists of nine β-strands from residues 400–495 , as shown in Figure 8A . All β-strands are antiparallel . Several loops between the β-strands ( i . e . , Lβ1β2 , Lβ2β3 , Lβ6β7 , and Lβ7β8 ) exhibit larger RMSD values compared with the β-strand regions in the structure , suggesting that these loops exhibit higher flexibility . The overall structure of the LysGH15 SH3b domain is very similar to the SH3b domain of the ALE-1 protein ( which is a peptidoglycan hydrolase produced by Staphylococcus capitis EPK1 that can specifically lyse S . aureus ) ( Figure 8B ) [24] . These proteins share approximately 47% sequence identity ( Figure 4C ) . A comparison of the structures of the two SH3b domains reveals a major difference in that the residues of ALE-1 corresponding to β7 and β8 of the LysGH15 SH3b domain adopt a loop conformation . The loop Lβ7β8 and strands β7 and β8 pack against β2 and β5 in the LysGH15 SH3b domain , whereas the corresponding region in the ALE-1 SH3b domain exists as a loop that points toward the outside of the protein . In the sequence alignment , two additional proline residues are present in the Lβ7β8 loop of the LysGH15 SH3b domain , and these proline residues participate in hydrophobic interactions with residues in β2 and β5 , which may explain this structural difference . Additionally , the construct used in the determination of the ALE-1 SH3b domain structure contained an N-terminal FLAG-tag , which occupies the region corresponding to the Lβ7β8 loop of the LysGH15 SH3b domain . In light of the sequence and structural similarities with the SH3b domain of ALE-1 , in addition to an identical target bacteria ( S . aureus ) , the peptide “AGGGGG” was used to perform NMR titrations of the 15N-labeled LysGH15 SH3b domain , as previously reported for the SH3b of ALE-1 [24] . Significant chemical shift perturbations ( CSPs ) in a fast-exchange manner ( on the NMR timescale ) upon peptide addition are observed in the 1H-15N HSQC spectra of the LysGH15 SH3b domain , as shown in Figure 8C and Figure S5A . The equilibrium dissociation constant KD is approximately 3 . 0 mM , as obtained by fitting the titration curve of the CSPs ( Figure S5B ) , which indicates that the binding is quite weak and may explain the inability to cocrystallize the ALE-1 SH3b domain with the polyglycine peptide in a previous study [24] . Residues exhibiting large CSP values are clearly clustered in the sequence , and mapping of the CSP results onto the structure of the LysGH15 SH3b domain reveal that these residues may interact with the peptide . The residues with significant CSPs are largely from β1 , β2 , β5 , loop Lβ1β2 , and Lβ3β4 , which form a groove for peptide binding ( Figure S4C and Figure S5C ) . The potential binding site is consistent with the polyglycine binding site of the ALE-1 SH3b domain that was proposed by previous docking and mutagenesis analysis [24] , [25] . However , S430 is the residue with the most significant CSP in the amidase domain of LysGH15 ( the corresponding residue in ALE-1 is G299 ) . To determine the effect of these residues on the binding , these residues were mutated to alanine in the individual SH3b domain of LysGH15 . As shown in Figure 8D , the mutations N404A , Y406A , G407A , or T408A in the SH3b domain significantly diminish the binding activity . Additionally , the mutations L433A and I454A also affect the binding of the SH3b domain . Although S430 exhibits a large CSP value in the NMR titrations , S430A does not substantially affect the activity of the SH3b domain , which indicates that the side chain of S430 may not be involved in the binding . The large chemical shift perturbation of S430 may be caused by the main chain of G429 that participates in the binding: the carbonyl of G429 points into the binding groove and may form a hydrogen bond with the HN of the peptide , which induces a large chemical shift of the S430 HN . Additionally , the effects of these residues on the lytic activity of the full-length LysGH15 were investigated . From Table 1 , the bactericidal activity of LysGH15 is reduced when these residues ( except with S430 ) are mutated to alanine .
In this study , we report that the lytic activity of the LysGH15 and its CHAP domain is critically dependent on a calcium ion . To confirm these findings , the structures of the individual CHAP , amidase-2 , and SH3b domains of LysGH15 were determined . This study represents the first report of the structure of a staphylococcal phage lysin . Interestingly , the structural studies reveal that the CHAP domain of LysGH15 contains a calcium-binding site that is located near the active site groove . This finding is unexpected because the calcium-binding site is not detectable through sequence-based searches alone [26] . To the best of our knowledge , LysGH15 is the first characterized lysin that contains a calcium-binding site . However , LysGH15 is not the first lysin to demonstrate calcium-dependent lytic activity . Although we have investigated the calcium-dependence of LysGH15 in our previous study [11] , the LysGH15 used was not pretreated with EDTA , which masked the detection of this important phenomenon . It has been reported that the LysK CHAP domain [27] , the staphylococcal phi11 lysin [28] , the streptococcal B30 lysin [29] , and the streptococcal Ply700 lysin [30] exhibit calcium-dependence . A classical EF-hand protein contains a helix–loop–helix Ca2+-binding motif . The “EF-hand-like” motif differs from the classical EF-hand as follows: ( i ) the length of the Ca2+-binding loop is shorter or longer than 12 residues and/or ( ii ) the secondary structure elements of the flanking regions are not two helices [26] . Although the calcium-binding loop of the LysGH15 CHAP domain is 12 residues in length ( the coordination residues lie at positions 1 , 3 , 5 , 7 , and 12 ) , the secondary structure elements surrounding the calcium-binding site of the CHAP domain only contain one helix , which is consistent with an “EF-hand-like” motif . Thus , the LysGH15 CHAP domain represents an “EF-hand-like” protein . As in the protective antigen from Bacillus anthracis ( PDB ID: 1ACC ) [31] and the dockerin from Clostridium thermocellum ( PDB ID: 1DAQ ) [32] , the calcium-binding site of the LysGH15 CHAP domain lacks the exiting helix , forming a “loop-F” pattern . Notably , the functions of these proteins available in the PDB that contain a calcium-binding site are completely unrelated to those of the LysGH15 CHAP domain . To date , the CHAP domain of LysGH15 is the first identified “EF-hand-like” protein that originates from a phage lysin . As in the thermolysin-like protease , the calcium ion plays an important role as a switch that modulates the protease between active and inactive states according to the biological demand [33] . Note that D45 , D47 , and D56 coordinate the calcium ion via their side chains , whereas Y49 and H51 coordinate the calcium ion via their main chains . Therefore , it is not surprising that the activity of the LysGH15 CHAP domain is retained upon mutation of Y49A and H51A . The reason for the calcium-dependence of LysGH15 is not clear . However , in light of its location near the active site groove and its significant influence on the lytic activity of the protein , there are two potential functions for this calcium ion: ( i ) it participates in the catalytic activity as part of the reaction; or ( ii ) it positions the key residues , particularly C54 , to form the appropriate conformation . The C54S/C54A mutation completely abolishes the lytic activity , which indicates that the sulfhydryl of C54 acts as a nucleophile and plays a critical role in the hydrolysis . This finding is consistent with the conclusion obtained from studies on the E . coli glutathionylspermidine ( GSP ) synthetase , which operates via a nucleophilic mechanism involving C59 as the catalytic nucleophile [34] . Some members ( 6 . 56% , 366/5579 ) in the Pfam sequence database ( http://pfam . sanger . ac . uk/family/PF05257 ) , which exhibit high sequence identity with the LysGH15 CHAP domain in the calcium-binding site ( particularly at positions 1 , 3 , and 12 ) and in the Cys-His-Glu-Asp proteolytic active site ( several sequences are provided in Figure 4D ) , are likely to also coordinate calcium , and their hydrolytic activity is expected to be dependent on calcium . Additionally , 36 . 04% ( 71/197 ) of the members in the family C51 of the MEROPS peptidase database ( http://merops . sanger . ac . uk/index . shtml ) also exhibit this sequence identity . As was observed for the inactive staphylococcal Φ11 lysin [15] , the streptococcal λSA2 lysin N-acetylglucosaminidase domain [35] , [36] , and LysK [37] , [38] , the LysGH15 amidase-2 domain alone is silent during activity analysis . Nevertheless , the LysGH15 amidase-2 domain is not entirely silent but additionally exhibits the ability to enhance the lytic activity of the CHAP domain . Therefore , the amidase-2 domain may cleave specialized substrates of the peptidoglycan , such as the bond between N-acetylmuramoyl-L-alanine [37] . The two domains may be able to simultaneously cleave the peptidoglycan between D-alanine and glycine , as well as Mur-NAc and L-alanine [37] . These large defects in the superstructure of the cell wall would result in rapid bacterial lysis [21] . The complex of AmiD with its substrate has been described , and the catalytic mechanism of AmiD has been elucidated [39] . Although the sequences of the LysGH15 amidase-2 domain and AmiD are divergent , they exhibit homologous structures and share conserved active site residues . In AmiD , E104 ( E282 in the LysGH15 amidase-2 ) and a zinc ion activate the water that is bound to the zinc ion , which favors the nucleophilic attack of the amide bond; the tetrahedral intermediate is stabilized by K159 ( this site is occupied by T330 in the LysGH15 amidase-2 domain ) . Mutational analysis demonstrates that E282 and T330 are critical for the activity of the LysGH15 amidase-2 domain . Thus , the amidase-2 domain of LysGH15 most likely possesses a similar catalytic mechanism as AmiD . Asparagine has been reported to be the only residue conserved among bacterial and eukaryotic amidases that participates in peptidoglycan binding [1] . Here , we find that the amidase-2 domain of LysGH15 from phage ( virus ) also contains this conserved residue ( i . e . , N275 ) at a corresponding position as other amidase members . These results provide further evidence for the hypothesis that “a common peptidoglycan binding mode is shared by all proteins with an N-acetylmuramyl-L-alanine amidase-like fold” [1] . Although the SH3b domain does not possess lytic activity [14] , in light of the large difference in the activity of full-length LysGH15 and its CHAP domain alone , the SH3b domain is expected to be necessary for LysGH15 to display high processive activity . The NMR titration indicates that the residues exhibiting large CSP values formed a deep and narrow groove . Additionally , most of these residues significantly affect the binding activity of the SH3b domain and the lytic activity of LysGH15 . We assume that the binding of the SH3b domain to its cognate receptor is dominant in localizing the catalytic domain to the cell wall . Once the catalytic domain is positioned close to the peptidoglycan layer , the local concentration of substrate is greatly enhanced , and significant catalysis may ensue [40] . Thus , both the binding activity of the SH3b domain and the catalytic activity of the CHAP contribute to the high processivity of LysGH15 .
The genes for the full-length LysGH15 and its three individual domains ( i . e . , CHAP , amidase-2 , and SH3b ) were amplified using corresponding primers that were designed based on the full-length lysGH15 gene ( GenBank: AY176327 ) and were synthesized by Sangon Biotech ( Shanghai ) Co . , Ltd . The coding regions for the CHAP ( residues 1–165 ) , amidase-2 ( residues 165–403 ) , and SH3b ( residues 368–495 ) domains were cloned into the pMCSG7 vector as previously reported [41] . The full-length lysGH15 gene was subcloned into the pET-26b vector . Mutations were designed based on these constructs and were generated using the QuikChange Site-Directed Mutagenesis Kit following the manufacturer’s instructions ( Stratagene ) . All of the recombinant plasmids were sequenced to verify the sequence . The plasmids harboring the target gene , which encoded 6× His-tagged proteins , were transformed into E . coli BL21 ( DE3 ) ( Tiangen Biotechnology ) . The cells were grown in Luria-Bertani ( LB ) medium at 37°C until the OD600 reached 0 . 8 . The culture was then induced with 0 . 2 mM isopropyl-β-D-thiogalactoside ( IPTG ) for 20 h at 16°C . Cells were harvested by centrifugation at 4 , 670×g for 30 min and were resuspended in phosphate-buffered saline ( PBS; 137 mM NaCl , 2 . 7 mM KCl , 50 mM Na2HPO4 , and 10 mM KH2PO4 , pH 7 . 4 ) . After lysis by sonication , the cell debris was removed by centrifugation at 38 , 900×g for 30 min . The supernatant was applied to a nickel-nitrilotriacetic acid ( Ni-NTA ) resin gravity column ( Qiagen ) that had been previously equilibrated with PBS . The column was washed using 100 ml of lysis buffer containing 20 mM imidazole , followed by a 50 mM imidazole wash . Finally , the protein was eluted with PBS containing 500 mM imidazole . After buffer exchange , the 6× His-tag was removed using tobacco etch virus ( TEV ) proteolysis ( except for full-length LysGH15 ) . Uncut protein was removed using a second Ni-affinity chromatography step . The proteins without a His-tag were concentrated and applied to a Superdex G200 size-exclusion chromatography column ( Amersham ) that was preequilibrated with 20 mM Tris-HCl ( pH 7 . 5 ) and 150 mM NaCl ( 500 mM NaCl for full-length LysGH15 ) . For the SH3b domain , 40 mM Na3PO4 and 50 mM NaCl , pH 6 . 5 , were used . Fractions containing the purified target protein were pooled and stored at −80°C until further analysis . The E . coli BL21 ( DE3 ) strain that contained the pMCSG7-CHAP vector was grown in M9 medium containing glucose ( 0 . 2% M/V ) , MgSO4 ( 1 mM ) , and ampicillin ( 100 µg/ml ) at 37°C until the OD600 reached 0 . 8 . Subsequently , selenomethionine was added to the culture ( 50 µg/ml ) . The subsequent purification steps were similar to those used for the native protein . The plasmid pMCSG7-SH3b was transformed into E . coli BL21 ( DE3 ) . The cells were grown in M9 medium containing glucose ( 0 . 2% M/V ) , MgSO4 ( 1 mM ) , and ampicillin ( 100 µg/ml ) . 15N ammonium chloride and/or 13C glucose was used as the sole nitrogen and carbon sources , respectively , for isotope labeling . Labeled SH3b was purified using an identical procedure as that used for the native protein . All of the proteins were initially screened for crystallization using the hanging-drop vapor diffusion method and commercially available sparse matrix screens at 16°C . Crystals were obtained by mixing 1 µl of the protein solution with an equal volume of the reservoir solution and equilibrating the mixed drop against 300 µl of the reservoir solution . Crystals of CHAP and Se-Met-CHAP were grown in a solution containing 0 . 1 M Bis-Tris-propane , pH 7 . 5 , and 3 . 8 M sodium formate using 10–15 mg/ml of the proteins . The crystal of the amidase-2 domain was grown using 10–18 mg/ml of the proteins and a reservoir solution containing 0 . 1 M Tris-HCl , pH 9 . 0 , 0 . 2 M Li2SO4 , 30% ( wt/vol ) PEG 3 , 000 , and 0 . 1 M xylitol . For phase determination , crystals of amidase-2 were soaked for 5 min in crystallization solution supplemented with 10 mM potassium iodide ( KI ) prior to cryoprotection and freezing . Diffraction data for native amidase-2 domain crystals and crystals soaked with KI were collected at beamline BL5 . 0 . 1 ( Advanced Light Source , Lawrence Berkeley National Laboratory , USA ) . Otherwise , diffraction data for the native CHAP domain crystal and anomalous diffraction data for the selenomethionine CHAP domain crystal were collected at 100 K using an ADSC Q315 CCD detector at beamline BL17U1 of the Shanghai Synchrotron Radiation Facility ( SSRF ) . The data sets were indexed , integrated , and scaled using the HKL2000 software [42] . The initial phases were determined using the X2DF structure determination pipeline [43] , [44] and the Se-SAD method [45] , and the initial model was built by PHENIX AutoBuild [46] . The models were manually improved in Coot [47] . Refinement was alternately performed using REFMAC [48] and PHENIX Refine [46] . Statistics for the data collection and refinement are summarized in Table S1 . NMR samples of the SH3b domain ( labeled by 15N and/or 13C ) contained 0 . 02% ( w/v ) sodium 2 , 2-dimethylsilapentane-5-sulfonate ( DSS ) and 10% ( v/v ) 2H2O . All NMR experiments were performed at 298 K on an Agilent DD2 600 MHz NMR spectrometer that was equipped with a Z-gradient triple-resonance cryoprobe , as previously described with some modifications [49] . Two-dimensional 1H-15N and 1H-13C HSQC and , three-dimensional CBCA ( CO ) NH , HNCACB , HNCO , HN ( CA ) CO , HBHA ( CO ) NH , HCCH-TOCSY , and CCH-TOCSY experiments were performed for SH3b backbone and side chain assignments . Three-dimensional 1H-15N and 1H-13C NOESY-HSQC spectra with mixing times of 150 ms were collected to generate distance restraints . All data were processed using NMRPipe [50] and were analyzed using NMRViewJ [51] . Proton chemical shifts were referenced to the internal DSS , and 15N and 13C chemical shifts were referenced indirectly [52] . The structures of SH3b were initially calculated using the program CYANA [53] and were then refined using CNS [54] with manual assignments as well as semi-automated NOE assignments performed using SANE [55] . Backbone dihedral angle restraints that were obtained using TALOS-N [56] and hydrogen-bond restraints according to the regular secondary structure patterns were also incorporated into the structural refinement . From the 100 initial structures , 50 of the lowest energy conformers of SH3b were selected for water refinement using CNS and RECOORDScript [57] , and the 20 lowest energy conformers were selected to represent the final ensemble of structures for SH3b . The quality of the structures was analyzed using MOLMOL [58] and PROCHECK-NMR [59] . Statistics for the data collection and refinement are summarized in Table S2 . The CA-MRSA strain USA300-TCH1516 was obtained from the American Type Culture Collection ( ATCC ) and was used throughout the study . Staphylococcal lytic assays were performed using an overnight culture of USA300 grown at 37°C in tryptic soy broth ( TSB ) supplemented with 1% wt/vol yeast extract . Staphylococci were washed in PBS ( pH 7 . 4 ) and resuspended at an OD600 of approximately 1 . 0 . Bacteria were mixed with the wt or mutant proteins , and the OD600 was kinetically measured in a spectrophotometer for 120 min . All assays were performed in triplicate . Peptidoglycan was isolated from stationary phase cultures of USA300 , as previously described with some modifications [1] . For a quantitative analysis of lysis , purified peptidoglycan from USA300 was dissolved in 50 mM sodium phosphate buffer and was adjusted to an OD578 of approximately 0 . 6 . When the protein was processed using EDTA ( 1 mM ) , the excess EDTA was removed by dialysis . The peptidoglycan lytic activity was measured as the decrease in the OD578 for 60 min . The peptide “AGGGGG” was synthesized by Sangon Biotech ( Shanghai ) Co . , Ltd . A stock solution ( 100 mM ) of the peptide was prepared in a buffer that was identical to that used for the SH3b protein sample . The interaction between SH3b and the peptide was detected by monitoring the two-dimensional 1H-15N HSQC spectra of SH3b during the titration . The observed CSPs were calculated as previously described [49] using the following formula:where δHN and δN are the changes in the 1HN and 15N chemical shifts , respectively . The equilibrium dissociation constants ( KD ) were estimated by fitting the CSPs to the following equation:where CSPmax is the CSP at the theoretical saturated condition , which was also obtained from the fit; r is the molar ratio of the peptide to the protein; Cpro is the concentration of the initial protein solution; and Clig is the stock concentration of the peptide . The protocols of ELISA used here were similar to those previously described with some modifications [24] . Briefly , polystyrene enzyme immunoassay 96-well plates ( Nunc PolySorp; Thermo Fisher Scientific ) were incubated with 100 µl of sonicated bacterial peptidoglycans at a concentration of 15 µg/ml in PBS at 4°C overnight . After coating , the wells were washed three times with distilled water , and the plate was subsequently blocked with 1% bovine serum albumin in PBS at 4°C overnight . The wells were washed three times with distilled water , and 100 µl of 10 µg/ml protein diluted in PBS was added to the wells and incubated at 4°C for 1 h . After incubation with protein , the wells were washed three times with PBS containing 0 . 05% Tween 20 . Anti-SH3b serum ( 100 µl ) diluted in PBS containing 0 . 1% bovine serum albumin was added to the wells and incubated at 37°C for 1 h . After three washes with PBS-Tween 20 , 100 µl of diluted goat anti-rabbit IgG horseradish peroxidase conjugate was added and incubated for 1 h at 37°C . Unbound conjugate was removed by washing three times with PBS-Tween 20 . Subsequently , 100 µl of substrate ( 3 , 3' , 5 , 5'-tetramethylbenzidine solution ) was added , and the reaction was stopped by the addition of 100 µl of 1 M H2SO4 . The optical density was measured at 450 nm in a spectrophotometer . Sedimentation velocity experiments were performed using a Beckman XL-I analytical ultracentrifuge at 20°C as previously described [60] . Briefly , protein samples were diluted with buffer ( 20 mM Tris-HCl , pH 7 . 5 , and 150 mM NaCl ) to 400 µl at an absorbance at 280 nm of approximately 0 . 75 . The samples were loaded into a conventional double-sector quartz cell and mounted in a Beckman four-hole An-60 Ti rotor . The data were collected at 262 , 000×g ( at the cell center ) at a wavelength of 280 nm . Interference sedimentation coefficient distributions were calculated from the sedimentation velocity data using SEDFIT ( www . analyticalultracentrifugation . com ) . Inductively coupled plasma atomic emission spectrometry ( ICP-AES , Varian , VISTA-MPX ) was used for the metallic evaluation of protein samples at Tsinghua University , as previously described with modifications [61] . The conditions for ICP-AES analysis were as follows: RF power , 1 . 15 kW; plasma gas flow rate ( Ar ) , 15 l/min; nebulizer gas flow rate ( Ar ) , 0 . 75 l/min; auxiliary gas flow rate ( Ar ) , 1 . 5 l/min; and viewing height , 12 mm . This analysis was performed using three replicates . Measurements were conducted at 16°C using an ITC-200 microcalorimeter ( GE Healthcare ) as described previously [62] . The samples were buffered with 20 mM Tris buffer pH 7 . 5 containing 200 mM NaCl . To determine the calcium-binding affinity of the CHAP domain , 600 µM CaCl2 was stepwise injected into 50 µM the Ca2+-free CHAP protein sample . The data were analyzed with the MicroCal Origin software . CD spectra were acquired on a Chirascan CD Spectrometer ( Applied Photophysics ) according to the previous description [63] . Freshly prepared CHAP protein with or without Ca2+ was adjusted to 0 . 15 mg/ml in 20 mM Tris , pH 7 . 5 , 200 mM NaCl prior to the measurements . Wavelength spectra were recorded at 20°C using a 0 . 1-cm path length cuvette . Each scan was obtained by recording every 1 nm with a bandwidth of 1 nm between the wavelength ranges of 200–260 nm . Thermal shift assays were conducted using 0 . 2 mg/ml of the CHAP protein with or without Ca2+ in a buffer [20 mM Tris ( pH 7 . 5 ) , 200 mM NaCl] supplemented with a 1 , 000 dilution of SYPRO Orange dye ( Invitrogen ) as described previously [64] . The fluorescence signals as a function of temperature were recorded using a real-time PCR machine ( CFX96; Bio-Rad ) in the FRET mode , in which the fluorescence intensity was measured with excitation/emmission 450–490/560–580 nm . The temperature gradient was set in the range of 20–95°C with a ramp of 0 . 5°C over the course of 15 s . Data were analyzed with the differential scanning fluorimetry analysis tool ( Excel-based ) , and the Boltzmann model was used for plotting melting curves of CHAP proteins to obtain the midpoint of the thermal unfolding value for CHAP proteins using the curve-fitting software XL fit 5 ( ID Business Solutions Ltd . ) . The atomic coordinates and structure factors have been deposited in the Protein Data Bank ( PDB ) under accession codes 4OLK for the CHAP domain , 4OLS for the amidase-2 domain , and 2MK5 for the SH3b domain . The BMRB accession ID for the SH3b domain is 19752 . Statistical significance was determined using the unpaired two-tailed Student’s t-test at a level of significance of P<0 . 05 . | The staphylococcal phage lysin LysGH15 demonstrates great potential against methicillin-resistant Staphylococcus aureus ( MRSA ) . Here , we report that the lytic activity of LysGH15 and its CHAP domain is dependent on calcium ions . To elucidate the molecular mechanism , we determined the structures of three individual LysGH15 domains using X-ray crystallography or nuclear magnetic resonance ( NMR ) . The crystal structure unexpectedly reveals an “EF-hand-like” calcium-binding site near the Cys-His-Glu-Asn quartet active site groove in the LysGH15 CHAP domain . Furthermore , the calcium ion plays an important role as a switch that modulates the lytic activity of the CHAP domain . Additionally , structure-guided mutagenesis also confirms that both E282 and the zinc ion play an important role in maintaining the lytic activity of the LysGH15 amidase-2 domain . Moreover , the NMR structure and titration-guided mutagenesis identify residues in the LysGH15 SH3b domain that are involved in the interactions with the substrate . The structure of LysGH15 is the first determined lysin structure from a staphylococcal phage , and these results represent a pivotal step forward in understanding this type of lysin . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"bacteriology",
"gram",
"positive",
"bacteria",
"staphylococcus",
"medical",
"microbiology",
"microbial",
"pathogens",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"microbiology",
"bacterial",
"pathogens"
] | 2014 | Structural and Biochemical Characterization Reveals LysGH15 as an Unprecedented “EF-Hand-Like” Calcium-Binding Phage Lysin |
Buruli ulcer ( BU ) vaccine design faces similar challenges to those observed during development of prophylactic tuberculosis treatments . Multiple BU vaccine candidates , based upon Mycobacterium bovis BCG , altered Mycobacterium ulcerans ( MU ) cells , recombinant MU DNA , or MU protein prime-boosts , have shown promise by conferring transient protection to mice against the pathology of MU challenge . Recently , we have shown that a recombinant BCG vaccine expressing MU-Ag85A ( BCG MU-Ag85A ) displayed the highest level of protection to date , by significantly extending the survival time of MU challenged mice compared to BCG vaccination alone . Here we describe the generation , immunogenicity testing , and evaluation of protection conferred by a recombinant BCG strain which overexpresses a fusion of two alternative MU antigens , Ag85B and the MU ortholog of tuberculosis TB10 . 4 , EsxH . Vaccination with BCG MU-Ag85B-EsxH induces proliferation of Ag85 specific CD4+ T cells in greater numbers than BCG or BCG MU-Ag85A and produces IFNγ+ splenocytes responsive to whole MU and recombinant antigens . In addition , anti-Ag85A and Ag85B IgG humoral responses are significantly enhanced after administration of the fusion vaccine compared to BCG or BCG MU-Ag85A . Finally , mice challenged with MU following a single subcutaneous vaccination with BCG MU-Ag85B-EsxH display significantly less bacterial burden at 6 and 12 weeks post-infection , reduced histopathological tissue damage , and significantly longer survival times compared to vaccination with either BCG or BCG MU-Ag85A . These results further support the potential of BCG as a foundation for BU vaccine design , whereby discovery and recombinant expression of novel immunogenic antigens could lead to greater anti-MU efficacy using this highly safe and ubiquitous vaccine .
Subcutaneous skin infection by Mycobacterium ulcerans ( MU ) leads to a potentially disfiguring , necrotic condition known as Buruli ulcer ( BU ) [1] . What often begins as an indolent skin nodule or small lesion can ultimately progress to expanding ulcerations , body-wide scarring , loss of limbs or eyes , and osteomyelitis [2] . These infections disproportionately affect children and are largely endemic to Sub-Saharan Africa , Australia , and Japan , where the unconfirmed mode of transmission is thought to be dependent on exposure to contaminated wetland areas and insect vectors [3 , 4] . Treatment regimens include lengthy combination anti-mycobacterial therapies , however , lack of medical access , absence of rapid and accurate diagnostics , and the often misleading symptoms of BU frequently lead to significant delays in therapeutic action [5 , 6] . At the point of extensive tissue damage , surgical debridement and skin grafting is required , resulting in significant morbidity and social stigmatization [7 , 8] . Antibiotics can be effective against MU if administered at an early time point prior to ulceration , and side effects of treatment can include nephrotoxicity and hearing loss [9] . While there is increasing promise for less toxic antibiotic therapies , currently no prophylactic vaccine is available to prevent BU in the areas with greatest prevalence [10] . BU vaccine research strategies have largely focused on prime-boost regimens using recombinant DNA and MU proteins , however , the efficacy of these approaches has not surpassed the transient , cross-reactive protection observed during experimental vaccination with tuberculosis vaccine strain , Mycobacterium bovis bacillus Calmette- Guérin ( BCG ) [11–18] . BCG , the most ubiquitous World Health Organization-approved vaccine administered across the world , possesses a promising safety profile but low efficacy against pulmonary tuberculosis afflicting millions of people [19 , 20] . Experimental BCG vaccination has been studied using BU animal models and has been shown to confer protection by delaying ulceration after murine footpad challenge with MU [11] . While BCG vaccination extends the time to appreciable footpad swelling , protection ultimately wanes and animal euthanasia is required . Retrospective studies in humans also provide support for the potential use of BCG as a foundation for an effective BU vaccine . Patients previously vaccinated with BCG were shown to have delayed onset to ulceration after infection with MU , as well as significant protection against developing complications of MU infection , such as osteomyelitis [21–23] . These lines of evidence further support the potential of BCG as a foundation for BU vaccine design , whereby improvement of BCG immunogenicity could lead to greater efficacy using this highly safe and ubiquitous vaccine . BCG has previously been engineered to express various recombinant immunogenic antigens and protein fusions for use in TB vaccine development , with numerous observed in vivo effects [24–26] . Recombinant BCG vaccine strains which have been engineered to overexpress major antigenic secretory proteins , such as ESAT-6 , TB10 . 4 , CFP10 , heat shock proteins , and members of the mycolyl transferase complex Ag85A , Ag85B , and Ag85C , have displayed the greatest promise by increasing both humoral IgG antibody production and CD4 mediated Th1 responses against M . tuberculosis challenge [27–33] . Similar strategies have been investigated in application to BU vaccine design as well , with varying degrees of success . Priming with a DNA-based vaccine encoding multiple MU polyketide synthase modules and boosting with recombinant protein by Roupie et al . yielded differential levels of antigen-specific IgG responses , as well as IFNγ and IL-2 secretion upon recombinant MU antigen stimulation [12] . However , no improvement in protection over the level conferred by BCG vaccination was observed . Alternatively , an investigation by Tanghe et al . used a similar strategy that employed the plasmid-based expression of MU Ag85A as a prime followed by a recombinant protein boost [13 , 14] . This vaccination regimen produced appreciable antigen-specific immunogenicity which correlated with a level of protection against MU challenge that was similar to that achieved by BCG vaccination alone . We recently utilized a combination of strategies from the TB vaccine field , as well as those used by previous attempts to design anti-MU vaccines by engineering a quality controlled recombinant strain of BCG that overexpressed MU-Ag85A [15] . In this study , we showed that not only could this vaccine strain significantly induce proliferation of antigen-specific CD4+ T cells and increase IFNγ+ Th1 splenocytes responsive to whole MU and subcellular fractions , but subcutaneous priming also decreased MU burden , protected against mycolactone-induced pathology , and extended the lifespan of MU-challenged mice to significantly greater levels compared with BCG vaccination alone . Knowing that overexpression of one MU antigen by BCG could have these effects , we were subsequently interested in determining if alternative antigens or combinations of antigens could yield improvements on vaccine immunogenicity and efficacy . The immunodominant antigens , Ag85B and TB10 . 4 , are two such antigens that have been successfully used to augment the protective qualities of BCG against experimental tuberculosis in animal models [29 , 31 , 32 , 34–37] . These individual antigens , as well as fusion proteins combining various small antigens or important T cell epitopes from multiple antigens , have also been successfully expressed heterologously in BCG . Importantly , these constructs have been shown to initiate production of antigen-specific CD4+ T cell populations known to be vital in generating the same anti-mycobacterial Th1 responses hypothesized to play a role in containment of MU in humans . Due to the encouraging results from our previous study involving BCG expression of Ag85A and the large body of evidence supporting the usefulness of Ag85B and TB10 . 4 antigens against other mycobacterial diseases , we generated a vaccine strain of BCG which expressed a fusion protein combining MU-Ag85B and the TB10 . 4 homolog from M . ulcerans , MU-EsxH ( BCG MU-Ag85B-EsxH ) . Here we will show that , compared to BCG vaccination , a single subcutaneous dose of BCG MU-Ag85B-EsxH induced significantly enhanced antigen-specific humoral responses , CD4+ T cell proliferation , and Th1 splenocyte responses in mice . In addition , a single , un-boosted , subcutaneous dose of BCG MU-Ag85B-EsxH conferred significantly greater protection compared to BCG by reducing bacterial burden in MU challenged footpads , resisting pathologies associated with MU infection , and significantly lengthened the lifespan of MU-challenged mice . Importantly , these effects were statistically improved over those conferred by BCG MU-Ag85A , which was the first and only vaccine strain superior to BCG vaccination against BU in mice .
Female , 6–8 week old , C57BL/6 mice were obtained from Jackson Laboratories . These mice were 14–16 weeks old by time of footpad challenge with MU . Animal work was approved by the Duke University Institutional Animal Care and Use Committee ( IACUCU protocol A065-13-03 ) . IACUC protocols performed at Duke University adhered to the AAALAC , USDA , Guide for Care and Use of Laboratory Animals and Public Health Service Policy on Humane Care and Use of Laboratory Animals and Animal Welfare Act . All strains of Mycobacterium bovis BCG-Danish ( BCGD ) were cultured at 37°C on solid Difco Middlebrook 7H10 agar or in liquid Difco Middlebrook 7H9 media supplemented with 0 . 5% glycerol , oleic-albumin-dextrose-catalase ( OADC ) , and 0 . 05% tyloxapol . Selection of BCG transformants expressing MU Ag85B-EsxH was accomplished by adding 25 or 50 μg/ml hygromycin to liquid or solid media , respectively . Liquid cultures consisting of volumes less than 50 ml were shaken at 120 rpm and larger volumes were expanded to 250 ml or less in one liter bottles rotated at 6 rpm . High-volume vaccine accession lots were aliquoted into 1 ml cryovials and frozen at a concentration of OD600 1 ( ~108 CFU/ml ) . Virulent M . ulcerans 1615 was kindly provided by Dr . Pamela Small ( University of Tennessee ) and was cultured at 32°C in Middlebrook media as prepared for BCG . For purification of plasmid DNA and sequencing , DH5α Escherichia coli ( E . coli ) was grown on lysogeny broth ( LB ) agar plates or in LB supplemented with 50 μg/ml hygromycin . The M . ulcerans Ag85B open reading frame and endogenous secretion signal were amplified from MU1615 genomic DNA and cloned into the mycobacterial vector pMV261 [38] , where antigen expression was mediated by the constitutive mycobacterial hsp60 promoter ( henceforth , pSL402 ) . Selection of plasmids was controlled through hygromycin resistance . The influenza hemagglutinin ( HA ) epitope was added to the C-terminus of the EsxH fusion . Electrocompetent BCG cells were prepared by centrifuging log phase culture ( OD600 0 . 6–0 . 8 ) at 3000 rpm for 10 minutes , followed by washing in a buffer of 10% glycerol and 0 . 05% tyloxapol . Mycobacteria electroporated with 0 . 5 μg plasmid DNA were recovered by shaking at 37°C in 1 ml Middlebrook 7H9 media overnight . A series of quality control characterizations was performed on vaccine accession lots by assessing expression of recombinant antigen , contamination , and plasmid DNA retention as previously described [39] . Bacterial lysates for immunoblot were prepared by centrifuging 10 ml of log-phase liquid culture at 3000 rpm for five minutes . After washing in phosphate buffer saline + 0 . 05% tyloxapol ( PBST ) , the final cell pellet was resuspended in 200 μl lysis buffer with glass beads and vortexed for three minutes . Lysates were clarified by collecting the supernatant after centrifugation for five minutes at 3000 rpm . Pre-cast SDS PAGE gels were loaded with 15 μl clarified lysate boiled in Laemmli buffer and run for one hour at 130V . Protein was then transferred to PVDF membranes for one hour at 30V . Blocking of membranes was performed by shaking in 5% fat free milk in TBS with 0 . 1% tween ( TBST ) for one hour at room temperature . For detection of the HA epitope , mouse anti-HA-HRP ( clone 3F10 , Roche ) was diluted in 5% milk-TBST ( 1:1000 ) and incubated room temperature for one hour . After washing in TBST , detection of proteins was performed using chemiluminescence ( Lumi-light , Roche ) . Plasmid DNA was purified using a modified Qiagen Miniprep protocol [39] . Isolated plasmids were heat-shocked into E . coli DH5α and plasmid DNA was re-purified from the resulting transformants . Correctly sized plasmid inserts were assessed by reaction with NdeI/EcoRV and analysis using gel electrophoresis . Plasmid inserts from 10 E . coli clones were sequenced and analyzed using Clone Manager software ( Sci-Ed ) . Vaccine contamination was assessed by spread plating 100 μl of thawed accession lot material on chocolate agar and examining for growth following a two week incubation at 37°C . For protection studies , C57BL/6 mice were subcutaneously vaccinated by injection of 100 μl ( ~107 cells ) from an accession lot vial of empty-vector BCG ( pHA ) , recombinant BCG MU-Ag85A , or BCG MU-Ag85B-EsxH into the scruff of the neck . Eight weeks post-vaccination , the mice were challenged intradermally with 105 M . ulcerans 1615 ( MU1615 ) via the footpad . MU1615 challenge inocula were consistently accessed from the same accession lot frozen at -80°C and were tested for virulence by pathology in mouse footpad models . The width and height of footpad swelling from infected mice were measured at two to three week intervals using digital calipers . To reduce animal suffering and in compliance with IACUC protocol , infected mice were euthanized once footpad swelling height exceeded 4 . 5 mm , prior to any visible ulceration . Quantitation of Ag85-specific CD4+ T cell levels was performed by flow cytometric analysis of MHCII tetramer staining . Mice were inoculated subcutaneously or intravenously ( via the retro-orbital route ) with 100 μl ( ~107 cells ) of a freshly thawed vaccine accession lot vial . Weekly blood samples were collected retro-orbitally followed by isolation of peripheral blood mononuclear cells ( PBMCs ) by gradient centrifugation using Lympholyte M ( Cedarlane ) . PBMC buffy coats were collected , washed in 10 ml phosphate buffered saline ( PBS ) , and were resuspended in 2 ml of ACK lysis buffer to remove erythrocytes . After centrifugation at 2000 rpm , PBMC pellets were washed with PBS and stained with APC-conjugated M . tuberculosis Ag85B-MHCII tetramer ( 1:500 , NIH Tetramer Core Facility ) in flow buffer ( 2% fetal bovine serum in PBS ) . The 15 amino acid epitope , FQDAYNAAGGHNAVF , was recognized by this tetramer and shares high sequence homology with MU Ag85 . PBMCs were stained with tetramer for 30 minutes at 37°C and then stained with FITC-conjugated anti-mouse CD4 ( 1:500 , clone GK-1 . 5 , Biolegend ) and PE-Cy5 anti-mouse CD8 ( 1:200 , clone 53–6 . 7 , Biolegend ) for 30 minutes on ice . The PBMCs were then washed in with flow buffer , centrifuged for 2000 rpm for five minutes , and were then resuspended in 4% paraformaldehyde . Following fixation cells underwent flow cytometric analysis using a Becton Dickinson ( BD ) LSRII and FlowJo software ( Tree Star Inc . ) . To stain effector and central memory cells , the tetramer protocol was followed by additional antibody incubation steps with APC-Cy7 anti-mouse CD4 ( clone GK-1 . 5 ) , FITC anti-mouse CD62L ( clone MEL-14 , BD Pharmingen ) , and PE-Cy7 anti-mouse CD44 ( clone IM7 , BD Pharmingen ) . MU bacilli present in infected footpads were quantified using a method previously described [15] . At 6 and 12 weeks post-challenge , footpads infected with MU1615 were removed from euthanized mice and for disinfection and dissection . The footpads were disinfected by five-minute contact with 70% ethanol , followed by three washes in PBST . The challenged footpads were then minced and crushed by mortar and pestle . Resulting homogenates were subjected to N-acetyl-L-cysteine ( NALC ) /NaOH treatment by adding a mixture ( 50:50 ) of 4% NaOH and 2 . 9% sodium citrate + 1% NALC for 20 minutes . Homogenates were pelleted at 3000 rpm for five minutes , washed with PBST , and larger particulates were removed by passage through a 40 μm mesh . 5 μl of filtrate was then evenly distributed within a 0 . 8 cm2 circle upon glass microscope slides . Heat-fixed smears were stained with auramine-rhodamine ( BD Biosciences ) , destained with acid alcohol , and counterstained with potassium permanganate . Stained slides were viewed under 100x oil immersion using a Nikon X microscope . Acid-fast bacilli ( AFB ) were counted in four random fields of view ( FOV ) per animal ( 16 images total per vaccinated group ) . Total AFB calculations were performed by multiplying cell counts by the number of 0 . 038 mm2 FOVs in each marked smear area per microliter of applied filtrate . ELISPOT plates ( PVDF , 96 well ) were equilibrated with 70% ethanol , washed with PBS , and were coated overnight with anti-mouse IFNγ antibody ( 1 μg/ml , clone AN18 , Mabtech ) . Necropsies were performed on mice eight weeks following intravenous vaccination to harvest splenocytes for growth in RPMI complete media ( RPMI with L-glutamine and 10% fetal bovine serum ) . 96 well plates were blocked using RPMI media and 6 x 105 splenocytes were combined in each well with varied agonists: MU-Ag85A peptide ( 100 μg/ml , FQAAYNAAGGHNAVWNFDDN ) , MU-Ag85B peptide ( 100 μg/ml , FQDAYNAAGGHNAVFNFNDN ) , heat killed MU ( HKMU , 1 mg/ml ) , or whole cell lysate prepared from log phase MU liquid culture . Splenocytes were stimulated for 16 hours at 37°C . The plates were then washed with PBS + 0 . 05% tween 20 followed by a two hour incubation with secondary anti-mouse IFNγ antibody ( 1:1000 , clone R46A2 , Mabtech ) at 37°C . After a wash and three hour room temperature incubation with VectaStain avidin peroxidase complex ( Vector Labs ) , plates were incubated with 3-amino-9-ethylcarbazole substrate for five minutes . The reaction was ceased by submersion of plates in deionized water . Spots were visualized and quantified using a CTL Immunospot plate reader . High-binding , 384-well plates ( Corning ) were coated overnight at 4°C with 100 ng/ml recombinant M . tuberculosis Ag85A , Ag85B , or purified Ag85 complex ( BEI resources , NIAID , NIH ) diluted in 0 . 1 M sodium bicarbonate . Wells were then washed once , blocked with 40 μl blocking buffer ( 4% whey protein , 15% goat serum , 0 . 5% tween 20 , and 0 . 05% sodium azide in PBS ) for one hour at room temperature , and then washed again . After blocking , of a 1:100 dilution of intravenously vaccinated mouse serum in blocking buffer was added for 2 hours at room temperature . Plates were washed four more times and 15 μl of a 1:1000 dilution of goat anti-mouse IgG ( Southern Biotech 1030–05 ) was added for one hour at room temperature . After four further washes , 20 μl of tetramethylbenzidine substrate was added per well for up to 15 minutes . The reaction was stopped by addition of 20 μl 0 . 33 N HCL solution and absorbance was read at 450 nm .
We have previously described the generation of quality controlled accession lots of recombinant BCG expressing antigens of interest , particularly MU-Ag85A [39] . In order to generate a vaccine strain of BCG which expressed an in-frame fusion between MU-Ag85B and the M . tuberculosis TB10 . 4 homolog , MU-EsxH , electrocompetent BCG was transformed with pSL402 ( Fig 1A ) . The pSL402 replicating plasmid controls transcription of MU-Ag85B-EsxH containing a C-terminal fusion to the influenza hemagglutinin epitope using the constitutive mycobacterial hsp60 promoter . Plasmid replication was regulated by the mycobacterial oriM and by oriE in E . coli shuttle strains . Selection of bacterial transformants utilized plasmid-encoded resistance to hygromycin . Fig 1B displays expression patterns of the MU-Ag85B-EsxH fusion protein in whole cell lysates from 3 randomly picked BCG transformants . A single transformant which strongly expressed the fusion antigen was selected to produce a large-volume vaccine accession lot upon which a quality control panel was employed to further characterize expression of recombinant antigen , purity of the vaccine lot , and integrity of the recombinant plasmid sequence [39] . Anti-mycobacterial immunity is largely governed by responses from CD4+ T helper cells [40 , 41] . In order to determine if antigen-specific adaptive immune responses could be generated by vaccination with BCG-MU-Ag85B-EsxH , C57BL/6 mice were either subcutaneously or intravenously or primed with 107 bacilli from thawed quality controlled vaccine lots prepared as previously described [39] . At weekly intervals , retro-orbital blood samples were collected to isolate peripheral blood mononuclear cells . Flow cytometric analysis of staining by MHCII tetramer was subsequently used to quantify the percentage of CD4+ T cells which recognized the Ag85 epitope , FQDAYNAAGGHNAVF . Fig 2A and 2B show the levels of antigen-specific T helper cells induced by intravenous or subcutaneous vaccination , respectively . Responses from BCG MU-Ag85B-EsxH were compared to those from mice vaccinated with BCG containing an empty expression vector , BCG MU-Ag85A , and unprimed mice . While low levels of Ag85-specific T cells were induced in response to endogenously expressed antigen in BCG , a significantly greater number of T cells was produced upon vaccination with BCG overexpressing Ag85A or Ag85B-EsxH . The vaccination route did differentially affect the speed and amplitude with which peak responses were reached . Intravenous inoculation induced quicker , larger , and more prolonged T cell responses compared to the subcutaneous route; an 8% Ag85-specific T helper cell population was reached at 3 weeks after the intravenous injection with BCG MU-Ag85B-EsxH compared to a 1 . 75% peak response post-subcutaneous vaccination . Greater statistical significance was also achieved between BCG Ag85B-EsxH versus empty-vector BCG compared to the enhancement of BCG MU-Ag85A responses during the intravenous injection at weeks 2 and 3 post-vaccination ( p<0 . 05 , and p<0 . 001 , respectively ) . Interestingly , the peak response to BCG MU-Ag85A reached a maximum of 5% compared to the 8% peak following vaccination with BCG MU-Ag85B-EsxH . However , subcutaneous vaccination resulted in highly similar T cell proliferative responses between BCG MU-Ag85A and BCG MU-Ag85B-EsxH , both of which were significantly higher than those induced by empty-vector BCG ( p<0 . 01 , p<0 . 002 ) . Additionally , we attempted to use MHCII tetramer staining to detect T cell populations capable of recognizing the M . tuberculosis TB10 . 4 epitope , SSTHEANTMAMMARDT ( data not shown ) . However , no tetramer positive populations were detected , which may be due to the presence of two amino acid substitutions at this tetramer peptide position within the MU-EsxH sequence . Together , these data suggest that like BCG MU-Ag85A , BCG MU-Ag85B-EsxH is capable of generating helper T cell populations responsive to Ag85 , an immunodominant antigen previously demonstrated to play a role in vaccine-mediated protection against MU . Evidence from previous studies has highlighted the importance of CD4+ memory T cell populations in establishing greater efficacy for mycobacterial vaccines [42] . To quantify the levels of memory T cells produced by vaccination with the recombinant BCG strains , C57BL/6 mice were intravenously primed with 107 bacilli and , four weeks later , peripheral lymphocyte staining was performed for CD4 and the CD62L and CD44 memory T cell markers . While the absolute numbers of CD4+ T cells rose in all vaccinated groups regardless of BCG strain , vaccination with BCG MU-Ag85B-EsxH induced significantly higher levels compared to BCG ( Fig 3A , p<0 . 05 ) . Of those populations , naïve T cells were significantly reduced upon vaccination with either BCG MU-Ag85A or BCG MU-Ag85B-EsxH ( Fig 3B , p<0 . 05 ) . Interestingly , both Ag85-specific CD4+ effector memory and central memory T cell populations were significantly higher upon vaccination with either BCG MU-Ag85A or BCG MU-Ag85B-EsxH compared to BCG alone ( Fig 3C and 3D , p<0 . 05 ) . These data suggest that BCG MU-Ag85B-EsxH could also be useful in establishing T cell memory reservoirs capable of recognizing an MU antigen known to be immunoprotective and possibly representing sources of anti-mycobacterial IFNγ and IL-2 [43] . The requirement of IFNγ-yielding Th1 responses for generating efficacious anti-mycobacterial immunity has been characterized by several studies [44] . To determine if production of antigen-specific T cells following vaccination with recombinant BCG strains could generate such responses , C57BL/6 mice were primed with empty-vector BCG , BCG MU-Ag85A , or BCG MU-Ag85B-EsxH . Eight weeks post-vaccination , mice were euthanized and harvested splenocytes were stimulated with various MU antigens: MU-Ag85A peptide , MU-Ag85B peptide , heat-killed MU1615 ( HKMU ) , or MU whole cell lysate . Quantification of IFNγ-producing splenocytes was performed by enzyme-linked immunospot ( ELISPOT ) and the numbers of IFNγ+ spot-forming units ( SFU ) detected after 24 hours of agonist stimulation were calculated ( Fig 4 ) . Vaccination with both BCG MU-Ag85A and BCG MU-Ag85B-EsxH yielded significantly increased IFNγ+ splenocytes compared to BCG alone when stimulated with all MU antigens ( p<0 . 05 ) . The greatest responses were detected upon stimulation with whole heat-killed MU , whereby priming with BCG MU-Ag85A and BCG MU-Ag85B-EsxH increased SFU over BCG pHA by 2 . 8-fold and 3 . 3-fold , respectively . Notably , cytokine-secreting splenocyte numbers trended higher in the BCG MU-Ag85B-EsxH-vaccinated mice compared to BCG MU-Ag85A following stimulation with MU-Ag85B peptide ( >2-fold increase ) and whole heat killed MU ( 1 . 2-fold increase ) . Four separate M . tuberculosis TB10 . 4 peptides were also used to stimulate splenocytes , but no appreciable IFNγ+ populations were detected . Interestingly , the response of BCG MU-Ag85B-EsxH vaccinated mice to HKMU was over double that of the SFU generated by stimulation with MU-Ag85B peptide alone . These data suggest that the BCG MU-Ag85-EsxH vaccine may increase responsiveness of functional Th1 cells to additional Ag85B peptides or to other antigens expressed by M . ulcerans cells . Humoral responses to mycobacterial infection are becoming increasingly recognized in adaptive defense and as potential therapeutics [45–47] . To determine if use of recombinant BCG could induce antigen specific antibody responses in vivo , C57Bl/6 mice were vaccinated with empty-vector BCG , BCG MU-Ag85A , or BCG MU-Ag85B-EsxH and peripheral blood was collected biweekly for 6 weeks . Sera were subsequently isolated and tested for antibodies specific to immunogenic Ag85 proteins by IgG ELISA . Fig 5A , 5B and 5C display time course antibody responses to recombinant M . tuberculosis Ag85A , recombinant Ag85B , and purified Ag85 complex , respectively . At 2 weeks post-vaccination , all IgG responses were low in mice vaccinated with BCG or BCG MU-Ag85A; however , BCG MU-Ag85B-EsxH induced high levels of anti-Ag85A , Ag85B , and Ag85 complex IgG . Over the course of 6 weeks , antibody kinetics varied depending on antigen and vaccine group . Antibodies produced by BCG alone were consistently lower than those elicited by recombinant strains but did increase over time . Ag85A IgG responses from BCG MU-Ag85A-vaccinated mice continued to rise over time compared to the BCG MU-Ag85B-EsxH response which began high and slightly decreased over 6 weeks . A similar trend was observed for BCG MU-Ag85A induced anti-Ag85B responses , however , BCG MU-Ag85B-EsxH vaccination yielded a very high response which did not decline over this time course . Finally , IgG antibody induction against purified Ag85 complex peaked for BCG MU-Ag85A-vaccinated mice at 4 weeks but began to decline by 6 weeks . Conversely , anti-Ag85 complex responses induced by BCG MU-Ag85B-EsxH immediately began high and continued to climb by week 6 . Compared to empty-vector BCG , BCG MU-Ag85A vaccination did not generate statistically significantly higher IgG responses for any antigens except to rAg85A at week 6 . However , BCG MU-Ag85B antibody induction was statistically significantly higher ( p<0 . 05 ) over empty-vector BCG for both rAg85A and rAg85B for all time points except anti-rAg85A at week 6 . Additionally , recombinant M . tuberculosis TB10 . 4 was plated in the same ELISA format , however no IgG could be detected by ELISA . Together these data highlight the ability of BCG MU-Ag85B-EsxH to induce high levels of IgG reactive to multiple immunogenic Ag85 antigens , with a more rapid initial response compared to BCG or BCG MU-Ag85A . We previously demonstrated that priming with BCG MU-Ag85A could significantly extend the survival time of MU challenged mice compared to BCG vaccination alone [15] . Upon characterizing the immunogenic properties associated with BCG MU-Ag85B-EsxH , some of which displayed enhancement over BCG MU-Ag85A , we were further interested in determining the protection profiles in similarly challenged mice . C57Bl/6 mice were subcutaneously primed with 107 empty-vector BCG , BCG MU-Ag85A , or BCG MU-Ag85B-EsxH and , 8 weeks later , were intradermally challenged with 105 virulent MU1615 via the footpad . The width and height of challenged footpads were measured with digital calipers during the period of infection . If footpad swelling surpassed 4 . 5 mm in height , mice were euthanized to reduce suffering . Fig 6A shows the time to euthanasia for unprimed and vaccinated mice . As previously demonstrated , while BCG vaccination increased the mean survival time from 6 . 3 weeks for unprimed mice to 8 weeks , subcutaneous vaccination with BCG MU-A85gA significantly increased survival time over BCG alone to 17 . 4 weeks ( p<0 . 01 ) . Markedly however , a single subcutaneous dose of BCG MU-Ag85B-EsxH further significantly increased the survival time of MU-challenged mice over that of BCG MU-Ag85A to a mean of 29 . 4 weeks ( p<0 . 001 ) . Previous BU studies in mice have demonstrated a correlation between the degree to which infected footpads swell and the MU bacterial load present within challenged tissue [15 , 48] . To determine if the observed enhancement of survival associated with BCG MU-Ag85B-EsxH vaccination correlated with a reduction in MU burden , mice which had received vaccinations and challenged as above were euthanized for isolation of MU1615 . Infected footpads were dissected at 6 and 12 weeks post-challenge and persisting acid-fast MU in filtered footpad homogenates were stained with fluorescent auramine-rhodamine . The evaluation of bacterial load using microscopy was previously assessed to confirm similar colony forming unit ( CFU ) results were achieved compared to plate counting [15] . Fig 6B shows the mean acid-fast burden for unprimed mice or those primed with empty-vector BCG , BCG MU-Ag85A , or BCG MU-Ag85B-EsxH . At both 6 and 12 weeks post-challenge , all subcutaneous vaccinations resulted in a significant reduction of footpad bacterial burden compared to unprimed mice . However , priming with BCG MU-Ag85B-EsxH consistently achieved the greatest protection at both time points , conferring a significantly greater reduction in footpad bacterial replication by 1 . 5 log and 2 . 56 log at 6 and 12 weeks post-infection , respectively ( p<0 . 001 , p<0 . 05 ) . Importantly , protection conferred by BCG MU-Ag85B-EsxH was also superior when compared to burdens present in empty-vector BCG ( a 0 . 78 log and 1 . 7 log reduction at 6 and 12 weeks , respectively ) and BCG MU-Ag85A-vaccinated groups ( 0 . 84 log reduction at 12 weeks post-challenge ) . This potent inhibition of bacterial growth by BCG MU-Ag85B-EsxH vaccination , as well as the intermediate inhibition by BCG MU-Ag85A correlated well with the distinct ability of each vaccine to extend survival times in the footpad challenge model . Production of the cytotoxin , mycolactone , is known to contribute to the tissue destruction histologically observed at the foci of MU infection [49 , 50] . To determine if the reduction of in vivo bacterial burden by vaccination correlated with protection against tissue damage , footpads were collected for histopathological analysis 12 weeks post-infection . Fig 6C displays representative images of hematoxylin and eosin ( H&E ) stained tissue sections from unprimed , empty-vector BCG , BCG MU-Ag85A , or BCG MU-Ag85B-EsxH primed mice . Consistent losses of epidermal layers , as well as extensive areas of internal necrosis and infiltrates of inflammatory cells were observed in footpads of unprimed mice . While BCG vaccination had a reduction in epidermal loss , substantial edema replaced necrotic lesions found in unprimed animals . However these features were rare in BCG MU-Ag85A and BCG MU-Ag85B-EsxH primed mice where , at 12 weeks post-infection , full tissue integrity remained in most animals . To visualize the organization of persistent MU bacilli in vivo , Ziehl-Neelsen ( ZN ) staining was performed on tissue sections from challenged footpads . Fig 6D displays ZN staining of the contiguous tissue sections of the above-mentioned H&E tissue sections . Large masses of pink acid-fast bacilli ( AFB ) present in the extracellular milieu as well as granulomatous lesions could readily be observed in tissue sections from unprimed MU-challenged mice ( Fig 6D and 6E ) , while fewer and smaller groups of AFB were detected in empty-vector BCG groups ( Fig 6E ) . However , AFB could not readily be observed in footpad sections from the BCG MU-Ag85 and BCG MU-Ag85B-EsxH vaccinated groups , correlating with the lower overall bacterial burden present in these tissues . Overall these data suggest that the protective BCG MU-Ag85B-EsxH immune responses , characterized by enhanced proliferation of antigen-specific Th1 CD4+ T cell populations and potent antigen-specific humoral IgG induction , can contribute to a reduction in bacterial burden , inhibition of tissue destruction , and an overall greater lifespan for MU1615-infected mice , compared to BCG and BCG MU-Ag85A .
Buruli ulcer is a neglected tropical disease whose persistence continues to inflict severe patient morbidity in the absence of rapid diagnosis and treatment . Today’s diagnostics are limited to molecular techniques , microbial culture , or histopathological analyses not readily available in the most afflicted regions . Currently , the standard of care involves lengthy medical regimens , including rifampin and streptomycin treatment , both of which may confer issues of toxicity . Generation of a prophylactic vaccine against M . ulcerans infection could be incredibly invaluable , especially due to the disproportionately high incidence of BU in children , the threat of long-term disfigurement , and associated social stigma . Despite many efforts to develop an efficacious BU vaccine , new candidates have either not displayed immunity or have conferred limited and short-lived protection against experimental MU infection . However , we recently have shown that a single subcutaneous dose of quality controlled BCG-based vaccine overexpressing the MU mycolyl transferase antigen 85A could significantly decrease bacterial burden , pathology , and increase survival time following MU1615 footpad challenge [15] . These protective effects were significantly greater than those conferred by the previously most protective vaccine strain to date , M . bovis BCG . Interestingly , a subsequent study highlighted MU-Ag85A as conferring protection when expressed in the M . marinum genetic background as well [51] . With the successful progress of recombinant BCG and M . marinum expressing MU-Ag85A as BU vaccine candidates , we decided to further improve a candidate vaccine strain by expressing other known immunodominant antigens . To this end , a novel recombinant BCG strain expressing the MU-Ag85B-EsxH fusion protein was generated , frozen into quality controlled accession lots , and evaluated for immunogenicity and protection against experimental BU in the mouse model . Several aspects of immunogenicity were investigated in the present study , including the induction of MU antigen-specific CD4+ T cell and memory populations , the production of IFNγ-secreting cells responsive to stimulation by whole MU and cellular components , and the production of the antigen-specific humoral responses . Interestingly , both BCG MU-Ag85A and BCG MU-Ag85B-EsxH were equally more significantly immunogenic compared to BCG when examining adaptive T cell responses and memory populations . In contrast , BCG MU-Ag85B-EsxH vaccination was capable of inducing significantly greater IgG responses to both rAg85A and rAg85B compared to empty-vector BCG at multiple time points post-vaccination , while inoculation with BCG MU-Ag85A was not . Although the importance of humoral responses for immunity against MU is not well characterized , antibody mediated protection might be of major relevance against advanced stages of MU infection , where bacilli are predominantly found as extracellular clusters . Furthermore , previous studies have demonstrated the potential efficacy of antibody-based therapies against experimental M . tuberculosis infection [45–47] . In addition to an evaluation of immunogenicity for the BCG MU-Ag85B-EsxH vaccine , we were interested in characterizing any improvement in protection during in vivo MU challenge compared to empty-vector BCG or BCG MU-Ag85A . Vaccination with a single subcutaneous dose of BCG MU-Ag85B-EsxH significantly increased the lifespan of infected mice over that of unprimed mice or empty-vector BCG by 4 . 7-fold and 3 . 7-fold , respectively . Strikingly , expression of Ag85B-EsxH in BCG also conferred significantly greater protection compared to BCG MU-Ag85A by increasing the mean survival time 1 . 7-fold compared to this strain . This result was also associated with a significant decrease in MU burden at 6 and 12 weeks post-challenge , with BCG MU-Ag85B-EsxH reducing bacterial burden compared to empty-vector BCG and BCG MU-Ag85A by 55-fold and 7-fold at 12 weeks , respectively . Of note , we were unable to demonstrate antigen-specific immune responses to the MU-EsxH protein . Proper molecular mass and plasmid sequencing of the fusion protein were confirmed in the BCG MU-Ag85B-EsxH accession lot according to our previously published quality control protocol , suggesting that the assays chosen were either not sensitive enough to detect EsxH immune responses or could not detect responses to EsxH as a fusion to Ag85B . Alternatively , the immunogenicity reagents used may have not represented a high enough homology to the MU-EsxH protein sequence . The TB10 . 4 MHCII tetramer , TB10 . 4 ELISPOT peptides , as well as recombinant TB10 . 4 protein used in the IgG ELISA were of M . tuberculosis origin , which shares 84% amino acid sequence identity with MU-EsxH . These reagents were chosen because of their ready availability from BEI resources; however , further characterization of antigen-specific immunity to the TB10 . 4 homolog will require the use of EsxH-specific reagents . This does insert a degree of complexity when determining the degree of contribution to protection each antigen provided , or if either antigen alone could singularly be responsible for the increase in protection observed in the present study . This would require a comparative analysis with BCG MU-Ag85B or BCG-EsxH , which is planned for future investigations . Previous vaccine studies which utilized Ag85 members and TB10 . 4 in tuberculosis models may shed light on their relative contributions towards immunity against MU . A recombinant strain of BCG which overexpressed the M . tuberculosis Ag85B was previously shown by Horwitz et al . to significantly reduce bacterial burdens in both the lungs and spleens of challenged guinea pigs compared to either recombinant Ag85B protein or BCG vaccination alone ( [31 , 37] ) . Mu et al . generated recombinant adenovirus vaccines which expressed M . tuberculosis Ag85A or the fusion between Ag85A and TB10 . 4 . Following vaccination and subsequent challenge with H37Rv , bacterial burden in the lungs of the antigen fusion strain was significantly reduced compared to the monovalent adenovirus-Ag85A and about 10 fold reduced compared to BCG alone [52] . Similarly , purified recombinant M . tuberculosis Ag85B or TB10 . 4 used as monovalent protein vaccinations was shown by Dietrich et al . to decrease bacterial burden in the mouse model of tuberculosis; furthermore , vaccination using either a mixture of the two or a fusion of the antigens significantly further reduced bacterial loads [34] . Interestingly , the fusion strategy for two MU proteins may itself confer an immunological advantage over separately expressed antigens , as shown by Palendira et al . [53] . In this study , vaccination with BCG separately overexpressing both M . tuberculosis Ag85B and the immunodominant M . tuberculosis antigen , ESAT-6 , was not as efficacious as a BCG strain encoding a fusion between the two antigens in reducing bacterial burden . Future improvements in strategies for BU vaccine design could be made to increase the protective qualities demonstrated by rBCG . We have shown that expression of MU-Ag85A or MU-Ag85B-EsxH both confer protection against the effects of MU infection , suggesting that future studies may have success in designing BCG strains which express a combination of these antigens or novel immunodominant MU proteins . Indeed , Sun et al . previously demonstrated that a triple antigen encoding strain of BCG ( AERAS-402 ) which expressed M . tuberculosis Ag85A , Ag85B , and TB10 . 4 generated stronger immune responses and conferred significantly improved survival compared to BCG vaccinated mice when challenged with a hypervirulent M . tuberculosis strain [36] . In addition to improvements in priming strategies , boosting with recombinant protein or other recombinant mycobacterial or viral vectors may also enhance vaccine efficacy . We previously demonstrated that boosting rBCG with recombinant M . smegmatis significantly improved protection in the mouse model of BU [15] . Furthermore , human clinical trials have effectively used recombinant modified vaccinia virus Ankara ( MVA ) expressing Ag85A to boost BCG by inducing potent and durable Th1-type responses as well as high levels of antigen specific T cell populations [30] . This approach has also recently been successful in boosting the triple antigen AERAS-402 strain of BCG to generate antigen-specific CD8+ T cells which produced high levels of anti-mycobacterial IFN-γ , TNF-α and IL-2 [35] . Use of mycolactone negative mutants or other attenuated MU strains may represent an alternative approach as well . There is experimental evidence to support this strategy , whereby subcutaneous vaccination with the mycolactone negative strain , MU5114 , yielded a delay to footpad swelling similar to that induced by BCG vaccination [41] . Furthermore , species of mycobacteria possessing greater genetic homology to MU than BCG may represent a richer source of protective antigens and lack the potential pathological features of an MU-based vaccine . Indeed , we and others have previously demonstrated that subcutaneous vaccination with M . marinum was significantly better able to delay the pathology of MU infection versus BCG alone [51 , 54] . Further novel attenuated and immunogenic MU strains , or strains with high genetic similarity to MU , may hold potential as priming or boosting agents to recombinant BCG . The well-supported safety profile , an established global administration infrastructure , reasonable cost of production , and previously demonstrated protection support BCG as an ideal vehicle for BU vaccine design . Further discovery of MU-specific antigens could be an invaluable source of protective immunogens and , upon combined expression by BCG , may achieve complete protection against experimental MU infection . This , in addition to identification of cross-reactive antigens which confer the intrinsic protection observed with BCG , will be essential to the application of this ubiquitous vehicle as an efficacious and safe Buruli ulcer vaccine . | Mycobacterium ulcerans ( MU ) infection causes a highly disfiguring , necrotic skin disease known as Buruli ulcer ( BU ) . Antibiotic treatments have low efficacy if the infection is diagnosed after ulceration begins , leading to frequent dependence on surgical removal of infected tissues . A prophylactic vaccine for BU does not exist and several attempts to create an effective vaccine have shown limited success . We recently demonstrated that a recombinant strain of M . bovis BCG expressing the immunodominant MU-Ag85A conferred significantly enhanced protection against experimental BU compared to the standard BCG vaccine . Here we show that BCG expression of a fusion between two alternative MU antigens , Ag85B and EsxH , can promote antigen-specific T cell and humoral immune response capable of significantly improving survival and protection against BU pathology , compared to BCG MU-Ag85A alone . These results support the potential for using the highly safe and ubiquitous BCG vaccine as a platform for further BU vaccine development . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"immune",
"cells",
"immunology",
"tropical",
"diseases",
"vaccines",
"preventive",
"medicine",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"vaccination",
"and",
"immuniz... | 2016 | Overexpression of a Mycobacterium ulcerans Ag85B-EsxH Fusion Protein in Recombinant BCG Improves Experimental Buruli Ulcer Vaccine Efficacy |
Duplications of genes encoding highly connected and essential proteins are selected against in several species but not in human , where duplicated genes encode highly connected proteins . To understand when and how gene duplicability changed in evolution , we compare gene and network properties in four species ( Escherichia coli , yeast , fly , and human ) that are representative of the increase in evolutionary complexity , defined as progressive growth in the number of genes , cells , and cell types . We find that the origin and conservation of a gene significantly correlates with the properties of the encoded protein in the protein-protein interaction network . All four species preserve a core of singleton and central hubs that originated early in evolution , are highly conserved , and accomplish basic biological functions . Another group of hubs appeared in metazoans and duplicated in vertebrates , mostly through vertebrate-specific whole genome duplication . Such recent and duplicated hubs are frequently targets of microRNAs and show tissue-selective expression , suggesting that these are alternative mechanisms to control their dosage . Our study shows how networks modified during evolution and contributes to explaining the occurrence of somatic genetic diseases , such as cancer , in terms of network perturbations .
Gene duplicability defines the propensity to retain multiple copies of a gene and varies among species and gene categories . In yeast , singleton genes , i . e . single copy genes whose duplication is selected against , preferentially encode members of protein complexes [1] , highly connected [2] , [3] and essential [1] , [4] proteins . Similar relationships are maintained also in multicellular species such as worm and fly , where singleton genes encode highly connected [2] and essential [5] proteins . The strict retention of one single copy of these particular gene categories is a consequence of the fragility towards dosage modifications . Their duplication is deleterious because it interferes with essential cellular functions and with the fine-tuned equilibrium between formation and disruption of protein-protein interactions [6] , [7] . Recent studies showed that the duplicability of mammalian hubs and essential proteins is different from that of other species . Human hubs [8] , [9] and mouse essential proteins that are involved in development [5] , [8] , [10] are preferentially encoded by duplicated genes , while other categories of essential mouse genes can be both singletons and duplicated [5] . These differences between human , mouse and the other species suggest that gene duplicability underwent modifications during evolution , which are likely related with the extensive acquisition of novel genes in vertebrates . Through massive gene duplication followed by diversification of paralogs , vertebrates accommodated the expansion of gene families that are involved in regulation , signal transduction , protein transport , and protein modification [11] , [12] . In this context , it has been proposed that a higher connectivity may favor the functional diversification of paralogs , for example through tissue specialization [8] . However , a thorough analysis of which types of genes undergo modification of their duplicability during evolution and how this influences the network properties of the encoded proteins is still missing . The comparison of gene and network properties between species is the most straightforward approach to verify whether the modification of gene duplicability is indeed related to the expansion of the vertebrate gene repertoire . Despite the fact that current representations of protein interactomes are still incomplete [13] , [14] , [15] and may include a high fraction of false positives [16] , the recent completion of interaction screenings in several species finally allows comparative network analyses . For example , the comparison of human , fly , worm , and yeast networks showed that they maintain a similar structure despite the difference in size [17] , [18] . In addition , regardless of their connectivity , proteins that occupy central positions in the interactomes of Saccharomyces cerevisiae , Drosophila melanogaster and Caenorhabditis elegans are also essential and slow-evolving [18] . These studies demonstrate that the comparison of protein and gene properties in different species can be used to infer general evolutionary trends . To unravel when the differences between duplicability and network properties arose during evolution , we undertake a comparative analysis of genes and networks in four species , Escherichia coli , yeast , fly , and human . These species display different levels of complexity , defined as the number of genes , cells , and cell types [11] , and also high quality genomic and interaction data . We compare connectivity and centrality of all proteins with origin , conservation and duplicability of the corresponding genes . We identify a core of singleton hubs whose properties are maintained constant from prokaryotes to human , and another group of duplicated hubs that have emerged during the evolution of vertebrates . Our analysis provides evidence of how the hubs properties modified during evolution and helps in interpreting the occurrence of somatic genetic diseases that are typical of multicellularity , such as cancer , in terms of network perturbations . In particular , we find that cancer genes are representatives of the two groups of human hubs: one that originated early in evolution and is composed of singleton genes , and the other that appeared later and is enriched in duplicated genes . Functionally , these two groups correspond to caretakers and gatekeepers , suggesting that these two different ways to initiate tumorigenesis emerged at different times during evolution .
The purpose of our analysis is to compare gene origin , conservation , and duplicability with connectivity and centrality of the encoded proteins in E . coli , S . cerevisiae , D . melanogaster , and Homo sapiens . To this aim , we identify a reliable set of unique genes in each species ( Table 1 ) , and develop a four-step procedure to determine origin , conservation , and duplicability of these genes ( Figure 1 ) . First , we retrieve all clusters of orthologs with different inclusiveness that are associated with each gene ( Figure 1A ) using the EggNOG database [19] . Second , we associate all 373 species present in EggNOG to seven internal nodes of the tree of life that represent major transitions in evolution ( Figure 1B ) . These nodes include the last universal common ancestor ( LUCA ) , which defines the ancestral organism before the split between prokaryotes and eukaryotes , eukaryotes , opisthokonts , metazoans , vertebrates , and mammals . We also consider group-specific transitions such as primates for human , insects for fly , fungi for yeast and bacteria for E . coli . Third , we identify orthologs and paralogs of each gene in the highest possible number of internal nodes ( Figure 1C ) . Finally , we exploit the information collected in the first three steps to assign gene origin , conservation , and duplicability ( Figure 1D , E , F ) . Since we retrieve orthologs for all species stored in EggNOG , we can use this information to infer general trends on gene origin , conservation , and duplicability during evolution . We define the evolutionary origin of a gene as the deepest internal node of the tree of life where an ortholog can be found ( see Methods ) . Overall , we observe high variability in the gene origin between species ( Figure 2A , Table S1 ) . In accordance with previous reports [20] , about 60% of human genes have orthologs in prokaryotes and early eukaryotes and more than one fourth of human genes originated with vertebrates or later . Similar trends are confirmed in other vertebrates but not in invertebrates , which are in fact composed of a higher fraction of old genes ( Figure 2A , Table S1 ) . The substantial acquisition of vertebrate-specific genes is likely related with the two events of whole genome duplications that occurred in the early vertebrate genome [21] , [22] . To measure gene conservation , we count the internal nodes of the tree of life where the gene is lost since it appeared . With this measure of conservation , we do not estimate sequence divergence within a set of orthologous genes , but rather retention or loss of orthologs throughout evolution . Moreover , by counting the number of missing instead of retained nodes , we obtain estimates of conservation that are comparable between species and independent from the time of appearance of the gene . Indeed , zero always corresponds to maximum conservation , while conservation decreases progressively with the increase in the number of nodes where no orthologs can be found . Among eukaryotes , invertebrates show a lower fraction of highly conserved genes ( conservation 0 , 1 , 2 ) and a higher fraction of poorly conserved genes ( conservation 4 and 5 ) when compared to vertebrates and fungi ( Figure 2B , Table S1 ) . Coupled with the results of Figure 2A , this suggests that invertebrates retain a high fraction of ancient genes that are lost in other lineages . To identify duplicated and singleton genes , we check whether paralogs are present within the eukaryotic-specific clusters of orthologs ( KOGs ) for eukaryotes , and within the most inclusive clusters of orthologs ( COGs ) for prokaryotes . As expected [4] , gene duplicability increases with the increase in organismal complexity ( Figure 2C , Tables 1 and S1 ) . Around 65% of human genes are duplicated , and similar percentages are found in other metazoans with the exception of insects , which have less than 60% of duplicated genes , ( Figure 1C , Table S1 ) . This result , together with the high rate of DNA loss [23] and the low rate of fixed transposable elements [24] , confirms the compactness of the fly genome [25] . We rebuild the interactomes of the four species by combining all available primary interaction data from seven public resources ( see Methods ) . Given the poor overlap between these datasets , their integration considerably increases the total number of interactions ( Table S2 ) , and the resulting networks are the most complete , to our knowledge , representations of protein interactomes ( Table 2 ) . Since these resources also contain interaction data for other species , we rebuild the interactomes also for Mus musculus and C . elegans in the attempt of extending the analysis to other species . However , the resulting networks represent only around 10% and 20% of the mouse and worm proteins , respectively . Due to this high level of incompleteness , we decide not to include these species in the analysis . The networks of human , fly , yeast , and E . coli are all scale-free ( Figure S1 ) , although they differ in terms of completeness , number of interactions , and type of experimental support ( Tables 2 and S2 ) . Because of this heterogeneity , and to minimize the impact of false positives , we identify a ‘gold set’ of interactions that are supported either by single-gene experiments or by more than one high-throughput screening . The only networks that retain a substantial fraction of information are those of human and yeast ( Table 2 ) . We use these two gold sets to confirm the signal obtained from the analysis of the whole networks , thus excluding that it is affected by the experimental differences between species . Since the networks that we rebuild are considerably bigger than those used in previous studies , as a first analysis we check whether we observe the same relationships between duplicability and connectivity that have been reported in the literature . We verify that , overall , more connected and more central proteins are encoded by duplicated genes in human and by singleton genes in the other species , both in the whole networks and in the gold sets ( Figure S2 ) . Singleton proteins are more connected than duplicated proteins also in fly , thus suggesting that the modification of the relationships between duplicability and connectivity occurred after the divergence of vertebrates . In order to verify whether the time of origin of a gene affects the network properties of the encoded protein , we analyze connectivity and centrality of each protein in respect to the origin of the corresponding gene . For each species separately , we compare degree and betweenness of proteins that originated at a given evolutionary time with degree and betweenness of all proteins that originated earlier and later . In each species , we find that genes of a given age encode proteins that are significantly more connected and more central than younger proteins and less connected and less central than older proteins ( Figure 3A , Table S3 ) . This means that older proteins established more interactions and became more central during evolution . The general tendency is detectable in all four species and in the gold sets of human and yeast . The only exceptions are ancient fly genes and human genes that originated with metazoans . In fly , the unstable signal may be influenced by the high fraction of interactions detected via high-throughput experiments ( Table S2 ) , which are enriched in false positives . The higher connectivity of human proteins that originated in metazoans is instead due to the peculiar features of these genes , which will become more evident with the analysis of duplicability ( see below ) . Since there is high variability in the number of genes that originated at each evolutionary time , we check whether this could affect the results . To this aim , we compare connectivity and centrality between random sets of 500 proteins originated at a given time and random sets of 500 younger and older proteins . After repeating the random comparison 100 , 000 times , we derive the distributions of the differences of mean degree and betweenness and compute the corresponding z-score . This is defined as the fraction of random comparisons with a difference <0 and >0 when compared with younger and older proteins , respectively . The analysis of these distributions confirms that proteins with a given origin are generally more connected and more central than younger proteins and less connected and central than older proteins ( Figure S3A and Table S3 ) . We next verify whether also the conservation of a gene has an impact on the network properties of the encoded protein . We compare degree and betweenness of proteins with a given conservation with degree and betweenness of more and less conserved proteins . By comparing both the total distribution of degree and betweenness with the Wilcoxon test ( Figures 3B ) and random sets composed of an equal number of genes ( Figure S3B ) , we observe that conserved proteins are connected and central , while proteins with low degree and low betweenness are also poorly conserved in all species . Although with a lower statistical support , the general trend is overall confirmed also in the gold sets of human and yeast ( Table S3 ) . Our analyses show that genes that appeared early in evolution and that are well conserved encode highly connected and central proteins . Since the same trend is found independently in all four networks , it is likely that these genes constitute a core of ancestral and conserved orthologs , which maintain identical properties throughout evolution . Indeed we find that between 44 and 51% of singleton hubs that originated early in evolution in one of the four species have orthologs that are singleton hubs also in one of the other networks ( Table S4 ) . This is a remarkable result , considering the level of incompleteness of the four interactomes and the fact that they are assembled independently from each other . Since we find that connectivity and centrality of a protein depend on when the corresponding gene appeared in evolution , we wonder how the gene origin affects the network properties of singleton and duplicated proteins . We compare connectivity and centrality between singleton and duplicated proteins that originated at the same evolutionary time . We find that , among ancient genes ( i . e . genes originated with LUCA and in early eukaryotes ) , singletons encode more connected and more central proteins than duplicated genes ( Figure 4A , Table S5 ) . Surprisingly , this tendency is detectable in all four species , including human , despite the opposite general trend of the human network ( Figure S2 ) . The difference between human and the other species arises when younger genes are analyzed . Human duplicated genes that originated with metazoans encode more connected and more central proteins than singleton genes of comparable age ( Figure 4A ) . For connectivity , this tendency is detectable also for genes that appeared in vertebrates and in mammals , although with lower statistical support . Again , the trend is confirmed in the gold sets ( Figure 4A ) . According to our findings , all species from prokaryotes to vertebrates maintain a group of highly connected proteins , which are encoded by ancient , conserved , and singleton genes that are sensitive to dosage modification . Another group of human hubs emerged later in evolution , namely with metazoans and , to a lower extent , with vertebrates and mammals . These genes differ from ancient hubs because they can retain gene duplicates and are therefore robust towards gene duplication . Their high connectivity explains why human genes that originated in metazoans deviate from the common trend and are more connected and central than older genes ( Figures 3 and S3 ) . In fly , the network properties of duplicated proteins that originated with metazoans do not differ from those of singletons . Therefore , metazoan-specific genes became central hubs at least after speciation of insects . This once again confirms that the modification in the relationships between duplicability and connectivity occurred in the ancestor of vertebrates . According to the results of our analysis , human hubs can be divided into two groups depending on their origin and duplicability . To test whether this distinction also results in the accomplishment of different biological processes , we compare the functions of these two groups of hubs . In absence of a consensus definition [26] , we identify hubs as the top 25% most connected proteins of the network , This results in 2 , 573 human proteins with more than 12 interactions . The comparison between the two groups of hubs shows that they are indeed involved in different processes ( Figure 4B , Table S6 ) . Ancient singleton hubs are enriched in basic functions that are needed for the survival of the cell , such as cellular metabolism and transcription . Duplicated hubs that appeared recently in evolution are instead involved in regulatory functions that coordinate the organization of the multicellular organism ( Figure 4B ) . We also notice that the time of appearance of a gene affects its function more than the duplicability ( Figure 4B , Table S6 ) . Ancient and recent hubs are therefore representative subgroups of ancient and recent genes , respectively . Similar functional differences between ancient and recent genes have been reported in yeast , where ancestral genes are involved in transcription , replication , and other basic cellular processes , while genetic , transcriptional , and posttranslational regulation is associated with recently evolved genes [27] . To understand how duplicated hubs adapted to the dosage imbalance due to gene duplication , we check whether they are ohnologs , i . e . paralogs originated via whole genome duplication [28] , miRNA targets , and tissue-selective genes . These are three different ways of controlling gene dosage . The duplication of the entire genome maintains the dosage balance between interactors and allows the duplication of dosage-sensitive genes in yeast [29] and in vertebrates [30] . Similarly , miRNAs play a pervasive role in the post-transcriptional regulation of gene expression in higher eukaryotes , particularly in those biological processes that require a fine-tuned control of the gene dosage , such as signal transduction [31] . Finally , tissue selectivity represents yet another mechanism of gene dosage control because paralogs expressed in different tissues do not interfere with each other [32] , [33] . We find that the fraction of duplicated hubs that are also ohnologs , miRNA targets , and tissue selective genes is significantly higher than that of singleton hubs ( Figure 5A , 61 . 4% and 33 . 9% , respectively , p-value <2 . 2×10−16 , Fisher's exact test ) . This enrichment is mostly due to the large overlap between ohnologs and duplicated hubs ( Figure 5B ) . However , the same trend remains detectable when only miRNA targets ( Figure 5C ) , and tissue selective genes ( Figure 5D ) are considered separately . Within duplicated hubs , these types of dosage regulation act on genes that appeared in metazoans and vertebrates more frequently than on genes that appeared earlier ( Figure 5 ) . One example that explains the role of miRNAs in tuning the gene dosage of paralogs is represented by atrophins , a phylogenetically conserved family of transcriptional regulators that appeared in metazoans ( Atro ) and duplicated in vertebrates ( ATN1 and Rere , Figure 6 ) . Atrophins are broadly expressed particularly during development [34] , [35] , [36] , and their modification leads to neurodegenerative defects in fly [36] and in vertebrates [37] . The dosage of the fly atrophin gene Atro is under the tight control of the microRNA miR-8 [38] ( Figure 6 ) . The lack of miR-8 produces Atro overexpression and results in elevated apoptosis in the brain , behavioral defects and severe defects in animal survival [38] , [39] . Also reduced Atro expression causes impaired survival , indicating that the fine-tuning dosage of this gene is crucial for its activity [38] . The gene dosage balance of the two atrophin paralogs seems to be tightly regulated also in vertebrates . Indeed , the Rere protein is able to directly bind the other atrophin paralog ATN1 , which is responsible for the neurodegenerative disorder dentatorubral-pallidoluysian atrophy ( DRPLA ) [40] , and to induce its massive re-localization in the nucleus upon overexpression [41] . Due to this direct interaction , it has been speculated that the modifications of Rere gene dosage may have a role in the pathogenesis of DRPLA [42] . Interestingly , Rere , but not ATN1 , is the target for the counterparts of miR-8 , i . e . miR-200b and miR-429 ( Figure 6 ) , which may regulate its dosage in a similar way [38] . In this scenario , it is reasonable to support a possible role of miR-200b and miR-429 in regulating the dosage balance between the two vertebrate atrophin paralogs .
In this study we show that the evolutionary history of a gene affects its duplicability , as well as the centrality and the connectivity of the encoded protein in the corresponding interactome . These results offer novel insights into the reciprocal influences between gene and network modifications during evolution . In all species , the core of the network is composed of ancestral and singleton hubs that are highly conserved because they do not require further modifications . Genes that are progressively acquired during evolution instead encode less connected and less central proteins . This agrees with the observation that essential proteins occupy the center of the network [13] , while proteins that are under positive selection and undergo structural modifications are located at the network periphery [43] . The importance of the time of origin on the properties of a gene has been recently reported also in yeast where proteins that originated before the whole genome duplication are more connected and more central than younger proteins [44] . Intuitively , these results support the preferential attachment model of network evolution , in which the expansion of the network starts from an ancient core [45] and progresses through gene duplication and divergence [46] . However , our analysis also reveals that significant deviations from this model occur in correspondence of massive genome reorganizations , such as the whole genome duplications that occurred in vertebrates . Owing to such events , even genes that are sensitive to dosage modifications can tolerate duplications because the dosage balance with their interactors is preserved . Therefore , together with the increase in the number of protein coding genes , vertebrates also modified their interactomes and likely both events played a role in shaping their evolution . The rapid functional divergence of paralogs through massive neo- and sub-functionalization [47] , [48] could also explain the retention of paralogous hubs owing to the quickly diversification of their function . However , sub- and neo-functionalization play a role in the diversification of paralogs also in other species such as E . coli , yeast , and fly , where only singleton hubs are retained . Therefore , the time of origin , more than the functional divergence , influences the retention of duplicated hubs . Conceptually and functionally , the two evolutionary distinct groups of ancient and recent human hubs resemble ‘date’ and ‘party’ hubs that have been described in the yeast interactome [49] . Similarly to party hubs , ancestral and singleton human hubs are mainly involved in cellular and nucleic acid metabolism , while recent and duplicated human hubs act as regulators , mediators or adaptors , similarly to date hubs . The difference between yeast and human is again in the time of appearance of human duplicated hubs and in the fact that in yeast both groups are encoded by singleton genes . Moreover , in human the signal of high connectivity and centrality that derives from recent hubs is stronger than that from ancient hubs ( Figures 3 , S3 , 4 and Table S5 ) . This is consistent with previous findings of an overall enrichment of the human network in duplicated hubs [8] , [9] ( Figure S2 ) . There are several indications that , despite being robust towards gene duplication , recent hubs remain sensitive to gene dosage modifications . First , human duplicated hubs rapidly underwent alternative ways to control their dosage , for example through tissue-selective expression and miRNA regulation ( Figure 5 ) . Second , ohnologs do not undergo further small-scale duplications and copy number variations [30] . Finally , genes that carry disease-related germline mutations are depleted in hubs [50] and somatic mutations of hubs are often associated with cancer [9] , [51] , [52] . All together , these observations indicate that hub modifications are usually harmful , even independently from the individual gene function . This analysis also adds novel insights to our understanding of the network properties of cancer genes and to the importance of gene dosage in the development of cancer . We recently reported that cancer genes are overall enriched in singleton hubs [9] . However , when the same analysis is repeated taking into account the gene origin , also cancer genes , like other human hubs , can be divided into two groups ( Figure S4 ) . One group is composed of ancestral cancer genes that encode singleton hubs , while the other includes cancer genes that originated with metazoans and are enriched in duplicated hubs . These two groups of cancer genes broadly correspond to caretakers , i . e . genes involved in the repair of DNA and in the maintenance of genome stability , and gatekeepers , which instead appeared lately in evolution and accomplish functions related to signaling and growth [53] . Therefore , there are two ways of promoting cancer , one that deals with basic and ancestral functions , and the other that interferes with regulatory processes . In either case , tumorigenesis starts from the somatic perturbation of hubs , which represent components of the cellular network that are sensitive to modifications .
For the four species considered in the analysis ( H . sapiens , D . melanogaster , S . cerevisiae and E . coli ) , we only use the protein entries present in EggNOG v . 1 . 0 [19] that are associated with unique gene identifiers . As sources of unique genes we consider RefSeq v . 37 entries [54] for human; FlyBase FB2009_01 [55] for fly; SGD ( frozen at January 5th 2010 ) [56] for yeast; and EcoCyc v . 14 . 0 [57] for E . coli . We gather protein-protein interactions from the non-redundant integration of seven public resources: BioGRID v . 2 . 0 . 49 ( February 1st 2009 ) [58] , IntAct ( frozen at January 23rd 2009 ) [59] , MINT ( frozen at February 5th 2009 ) [60] , DIP ( frozen at January 26th 2009 ) [61] , DroID v . 4 . 0 ( July 2008 ) [62] , HPRD ( September 1st 2007 ) [63] , and a recent map of yeast interactions detected by yeast-two-hybrid [16] . We only consider primary data ( i . e . interactions directly detected in each of the species ) , and discard putative interactions inferred from orthology . We distinguish between two types of experimental evidence: 1 ) single-gene experiments , i . e . studies that report less than 100 interactions; and 2 ) high-throughput experiments associated with large-scale screenings . We derive a gold set of interactions that only includes data that are supported by single-gene experiments or by more than one high-throughput screening . For each protein in the four networks we compute degree and betweenness . Degree measures the connectivity of a protein inside the network and is calculated as the number of binary interactions . Betweenness is a measure of centrality and is related to the number of shortest paths that pass through a protein [64] . We identify seven internal nodes of the tree of life that correspond to major transitions in evolution ( LUCA , eukaryotes , opisthokonts , metazoans , vertebrates , mammals , and group-specific transition ) , and assign each of the 373 species present in EggNOG v . 1 . 0 to the most specific internal node , using the corresponding taxonomy ID . The four analyzed species are assigned to the corresponding group-specific transition ( primates , insects , fungi , bacteria ) , while the remaining 369 species are taken as representatives of the other major transitions . For example , we assign human to primates , other non-primate mammalian species ( i . e . mouse ) to mammals , non-mammalian vertebrate species ( i . e . fish ) to vertebrates , and so on . The group-specific nodes for the four species do not reflect comparable evolutionary transitions , and for human we are much more specific than with the other three species . This reflects the availability of species and orthology information in EggNOG . For example , in human we are able to discriminate between genes that originated in mammals and genes that originated with primates because in EggNOG there are three primates ( H . sapiens , Pan troglodytes and Macaca mulatta ) and five additional mammals ( Monodelphis domestica , Bos taurus , Canis familiaris , M . musculus and Rattus norvegicus ) . For fly , instead , the group-specific transition is insects , because only three insects have orthology information ( D . melanogaster , Apis mellifera and Anopheles gambiae ) . It should be noted that this different resolution of the group-specific nodes does not introduce any bias in the results because the fraction of group-specific genes is very low in all species . In addition , the number of genes that originated at a certain time in evolution does not affect genes that originated earlier or later . Finally , the general trend of origin , conservation , duplicability and network properties is detectable in all species , independently on the resolution of the group-specific transitions . Once species have been assigned to internal nodes , we assign each gene to clusters of orthologs with different levels of inclusiveness and check for the presence of orthologs in the seven internal nodes . For example , for human we check for the presence of non-primate orthologs in the mammalian clusters , of non-mammalian orthologs in the vertebrate clusters , of non-vertebrate orthologs in the metazoan clusters , and so on . We define the origin of each gene as the most ancient internal node where an ortholog can be found . For a small number of genes in each species ( 120 in human , 270 in fly and 5 in yeast ) we cannot assign a precise evolutionary origin , because no clusters that contain the gene include representative orthologs . These genes are excluded from further analysis . To measure conservation of a gene throughout evolution , we count the number of missing nodes , i . e . internal nodes of the tree of life where no orthologs of that gene can be found since it originated . By considering the same number of internal nodes ( seven ) for all species and by counting the number of lost instead of retained nodes , we gather an estimate of conservation that is comparable between species and independent from the origin of the gene . We consider a gene duplicated if there is at least one other gene of the same species ( i . e . at least one paralog ) within the eukaryotic-specific clusters ( KOGs ) for human , fly and yeast , and within the most inclusive clusters ( COGs ) for E . coli . If no paralogs can be detected , the gene is considered singleton . With this method , we do not date the time of gene duplication but rather gene duplicability , i . e . whether a gene underwent duplication and this duplication was retained at least once in evolution . For a total of 63 genes in human , fly , and yeast both KOG nor COG clusters are available , and we exclude these genes from further analysis . We group genes according to their evolutionary origin and compare the distributions of degree and betweenness with the corresponding distributions of younger and older proteins . In a similar way , we compare the distributions of degree and betweenness of proteins with a given conservation with those of more and less conserved proteins . All comparisons are made using the Wilcoxon test . In order to eliminate possible biases due to the different number of genes that originated at each evolutionary time , we apply a randomization test . In each species independently , we extract 500 random genes with a given origin and calculate the mean degree and betweenness of the corresponding proteins . We then compute the difference between these values and the corresponding mean degree and betweenness of 500 randomly picked younger proteins and older proteins , separately . In case a group includes less than 500 genes , also the other groups will contain the same number of genes ( i . e . since there are only 84 primate-specific genes , they are compared to 84 randomly selected younger or older genes ) . We repeat the random comparison 100 , 000 times and derive the distributions of the degree and betweenness differences between the proteins that originated at a certain evolutionary level and younger and older proteins . Finally , we calculate the z-score as the fraction of random comparisons with a difference <0 when comparing with younger proteins , and >0 when comparing with older proteins . Differences in the mean degree or betweenness <0 are associated with more connected or central proteins , while differences >0 to less connected or central proteins . We use a similar random test to compare degree and betweenness of proteins with a certain level of conservation with more and less conserved proteins . To visualize the results , we transform the p-values and z-scores into heatmaps . Red boxes are associated with significantly higher values of degree and betweenness , green boxes correspond to significantly lower values , and non-significant p-values are colored in black . To evaluate the effect of gene origin on duplicability , we compare degree and betweenness of duplicated and singleton proteins with the same age using the Wilcoxon test . Also in this case , we derive the heatmaps from the p-values . Red-colored boxes indicate that duplicated proteins are more connected or more central , green-colored boxes indicate that singleton proteins are more connected or more central , and black indicates no statistically significant difference between singleton and duplicated proteins . To perform the functional analysis we rely on the biological process branch of the gene ontology ( GO ) tree and compare GO terms present at levels 5 and 6 [65] . GO levels refer to the branching points of the tree , with level 1 corresponding to the root of the tree . Increase in levels numbers are associated with increased specificity in the functional description and to decreased number of described genes . Levels 5 and 6 represent a compromise to obtain a good resolution in functional description for a fair number of genes . We further group all terms at these two levels into 12 categories and perform three comparisons: ( 1 ) ancient singleton hubs and recent duplicated hubs; ( 2 ) genes that originated in LUCA and eukaryotes ( ancient ) and genes that originated in metazoans and vertebrates ( recent ) ; ( 3 ) singletons and duplicated genes . For each comparison , the functional enrichment is detected using Fisher's exact test and the resulting p-values are adjusted for the false discovery rate ( FDR ) using Benjamini-Hochberg method . From the list of 4 , 174 human ohnologs , i . e . paralogs originated via whole genome duplication [22] , we identify 3 , 867 genes in our dataset that duplicated through whole genome duplications . Of these , 3 , 618 are duplicated genes , while the remaining 249 singletons are likely false positives and thus discarded from further analysis . To derive a list of human genes that are targets of microRNAs , we use Tarbase v . 5 ( June 2008 ) [66] and miRecords v . 1 ( August 15 , 2008 ) [67] , which collect 1 , 051 and 1 , 311 experimental interactions , respectively . Starting from the interactions , we derive 986 human miRNA target genes from the two lists ( Table S7 ) . Of these , 952 genes are also present in our dataset of 18 , 074 unique human genes . We retrieve expression data for 13 , 787 unique Entrez genes in 36 [68] and in 73 [69] human normal tissues ( six tumoral tissues were excluded from the analysis to avoid that the deregulation of gene expression due to the disease condition could influence the analysis ) . We obtain a cumulative dataset of 4 , 988 tissue-selective genes , by considering only genes that are expressed in less than 25% of the analyzed tissues ( 8 and 17 in the two studies , respectively ) . Of these , 4 , 616 genes are also present in our list ( Table S8 ) . From the obtained lists of ohnologs , miRNA targets and tissue-selective genes , we extract the genes that encode duplicated and singleton hubs ( Table S9 ) . We then compare the corresponding fractions of singleton and duplicated hubs that are also ohnologs , miRNA targets and tissue-selective genes altogether and separately . All statistics are done using R version 2 . 10 . 1 . | Gene copy number is often tightly controlled because it directly affects the gene dosage . In several species , including yeast , worm , and fly , genes that have a single gene copy ( singleton genes ) encode proteins with several connections in the protein interaction network ( hubs ) as well as essential proteins . Surprisingly , in mouse and human essential proteins and hubs are encoded by genes with more than one copy in the genome ( duplicated genes ) . Here we show that these two distinct groups of hubs were acquired at different times during the evolution of protein interaction network and contribute in different ways to the cell life . Singleton hubs are ancestral genes that are conserved from prokaryotes to vertebrates and accomplish basic functions that deal with the cell survival . Duplicated hubs were acquired mostly within metazoans and duplicated through vertebrate-specific whole genome duplication . These genes are involved in processes that are crucial for the organization of multicellularity . Although duplicated , also recent hubs are subject to gene dosage control through microRNAs and tissue-selective expression . The clarification of how the protein interaction network evolves enables us to understand the adaptation to the progressive increase in complexity and to better characterize the genes involved in diseases such as cancer . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"systems",
"biology",
"genomics",
"cancer",
"genetics",
"genome",
"evolution",
"genetics",
"biology",
"computational",
"biology",
"comparative",
"genomics",
"genetics",
"of",
"disease",
"genetics",
"and",
"genomics"
] | 2011 | Modification of Gene Duplicability during the Evolution of Protein Interaction Network |
Wild and domesticated Atlantic salmon males display large variation for sea age at sexual maturation , which varies between 1–5 years . Previous studies have uncovered a genetic predisposition for variation of age at maturity with moderate heritability , thus suggesting a polygenic or complex nature of this trait . The aim of this study was to identify associated genetic loci , genes and ultimately specific sequence variants conferring sea age at maturity in salmon . We performed a genome wide association study ( GWAS ) using a pool sequencing approach ( 20 individuals per river and phenotype ) of male salmon returning to rivers as sexually mature either after one sea winter ( 2009 ) or three sea winters ( 2011 ) in six rivers in Norway . The study revealed one major selective sweep , which covered 76 significant SNPs in which 74 were found in a 370 kb region of chromosome 25 . Genotyping other smolt year classes of wild and domesticated salmon confirmed this finding . Genotyping domesticated fish narrowed the haplotype region to four SNPs covering 2386 bp , containing the vgll3 gene , including two missense mutations explaining 33–36% phenotypic variation . A single locus was found to have a highly significant role in governing sea age at maturation in this species . The SNPs identified may be both used as markers to guide breeding for late maturity in salmon aquaculture and in monitoring programs of wild salmon . Interestingly , a SNP in proximity of the VGLL3 gene in humans ( Homo sapiens ) , has previously been linked to age at puberty suggesting a conserved mechanism for timing of puberty in vertebrates .
Both wild and domesticated populations of Atlantic salmon ( Salmo salar L . ) show large phenotypic variation for sea age at sexual maturity [1] . Salmon males can stay in the sea 1–5 years before they initiate sexual maturation and return to their native river to spawn , while females usually return to the river after 1–3 years in the sea . In aquaculture the variation in age at sexual maturation is a remaining problem since precocious puberty in males results in negative effects on somatic growth , flesh quality , animal welfare and susceptibility to disease [1] . Early maturation in farmed salmon can also increase the risk of genetic introgression of escaped salmon in wild populations [2 , 3] , as maturing fish will have a higher likelihood of migrating to a nearby river to spawn . Immature fish on the other hand will more likely migrate to sea where mortality is high before reaching maturity [4] . Salmonids in general display moderately high heritability ( up to H2p = 0 . 39 ) for age at sexual maturation [5–9] and QTLs relating to this trait have been identified previously [10] . Also three recent papers used single nucleotide polymorphism ( SNP ) arrays to identify markers associated with sea age at puberty in an aquaculture strain using a low density SNP array [11 , 12] and regions under selection in wild populations using a high density SNP array [13] . These three reports revealed association of the trait to multiple loci but gave no clear answer regarding possible mechanisms , genes and genomic regions behind age at puberty . Previous studies screened the genome for loci under selection using a limited set of SNPs , which may exclude the causative variants [14] . The recent sequencing of the Atlantic salmon genome ( [15] , AKGD00000000 . 4 ) provides an opportunity for large-scale mapping and comparison of parallel sequencing reads on to the published genome assembly for the species , thereby enabling genome-wide detection not only of novel SNPs , but also small indels and structural variation [16] . Hence , the use of sequencing allows prediction of how genetic variants affect regulatory regions and genes , which may link traits with new biological mechanisms and provide opportunities for subsequent functional studies . This study aimed to elucidate the genes and genomic regions that regulate sea age at puberty in male Atlantic salmon . Males were chosen since the mechanisms of maturation may differ between the sexes , and because precocious maturation of males in aquaculture represents a significant challenge in production of this species . To investigate this trait , we performed genome resequencing of scale samples from sexually mature wild salmon from six rivers in Western Norway , returning either after one or three years at sea ( Fig 1 ) . Using this approach we identified a region on chromosome 25 ( Chr 25 ) harboring a dense set of significant SNPs in a stretch of 370 kb . These results were also confirmed in other year classes of wild salmon and in domesticated salmon that had been reared under controlled aquaculture conditions . In conclusion , we show for the first time the importance of one single genomic region in determining age at maturity in male salmon .
To find SNPs associated with age at maturation in salmon males , we sequenced 20 salmon per river and sea winter age ( 1SW and 3SW ) . This number of individuals in each pool has been shown to be sufficient to identify causative SNPs for a trait in Drosophila melanogaster [17] . Mapping our data yielded a 12 . 32X mean coverage ( 0 . 24 SE ) of unique mapped reads per river and sea age at puberty ( S1 Fig ) . This depth of coverage is similar to what has been used in other successful genome wide association studies ( GWAS ) by pool sequencing in vertebrates , including pig ( Sus scrofa ) and chicken ( Gallus gallus ) [18 , 19] . We have mapped the salmon sequences to the most recent salmon genome assembly ( AKGD00000000 . 4 ) . Within this assembly , 34% of the genome has not been assigned to chromosomes , probably due to a high number of repetitive sequences and the partially tetraploid nature of the genome [20] . This probably also explains why the unassigned part of the genome harbored only 1% of our uniquely mapped reads ( S1 Fig ) . SNP calling revealed altogether 4 , 326 , 591 SNPs in all sea ages and rivers , the data has been deposited at http://marineseq . imr . no/salmatsnp/ . Comparing 1SW and 3SW allele frequencies using the Cochran-Mantel-Haenszel ( CMH ) test in 4 , 326 , 591 SNPs revealed 138 SNPs that were significantly associated ( 0 . 1% FDR ) with sea age at puberty ( Fig 2A and S1 Table ) . Several single significant SNP associations with the phenotype were detected on chromosomes 1–7 , 9–24 and 27–29 , although these were not found to be among the most significant SNPs ( Fig 2A and S1 Table ) . None of the loci harboring single significant SNPs were further assessed as candidate loci since the power of sequencing pools increases with numbers of reads assessed . Although we cannot rule out true association of single SNPs with maturation we regard such signals as likely false positives . In a previous QTL study for precocious parr maturation the trait was shown to be linked to Chr 12 [21] . Chr 12 has also been associated with sea age at maturation in another study [22] . In a GWAS , using a 6 . 5 kb SNP chip , the trait of 1SW maturation or “grilsing” was found to be weakly linked to both Chr 12 and Chr 25 [11] . In our study , 74 of the 138 ( 48% ) SNPs associated significantly with the trait were located in a region on Chr 25 , covering ~370kb ( Fig 2B ) . From our data we conclude that in Western Norway a single selective sweep on Chr 25 has had a large effect on sea age at maturity while other regions in the genome might contribute to a lesser degree . This is in contrast to earlier reports showing a more polygenic nature of this trait , with contributions from several genomic regions [11 , 12 , 22] . A previous theoretical based model study also suggested that age at maturity could be regulated by a stable genetic polymorphism , in accordance with our current findings [23] . To verify the GWAS findings and to ascertain whether genotypes of single individuals for SNPs are associated with sea age at maturity we designed a Sequenom assay for 11 of the most significant SNPs in the selective sweep found in Chr 25 ( S2 Table ) . Genotyping of all 240 individuals included in the sequenced pools used in GWAS confirmed a strong association between allele frequencies and age at maturity ( S2 Fig ) . To characterize haplotypes using the 11 assayed SNPs , we performed a pairwise disequilibrium analysis on all samples that had been sequenced [24] . This analysis revealed two dissimilar haplogroups comprising 11 haplotypes in one block ( Fig 3A ) . One and five of these haplotypes showed significant association with maturing early and late , respectively . The significant 1SW haplotype explained 54% ( β-value -1 . 0 , p-value = 3 . 88e-40 ) of the phenotypic variance for this trait . The most significant 3SW haplotype explained 21% ( β-value 0 . 66 , p-value = 3 . 98e-13 ) of the variation in age at maturity , the other four 3 SW haplotypes explained the 1 . 9 , 2 . 2 , 3 . 6 and 3 . 7% , adding up to 32 . 4% of the variance of the age at maturity in 3SW haplotypes . The genotyping data clearly confirmed our findings from the pool re-sequencing and further supported that this locus exerted a large effect on the trait . In samples from the pool sequencing we identified haplotypes associated with sea age at maturity in the 2008 year class ( year of migration to sea ) . These fish have possibly been exposed to similar environmental conditions during their early stay in the sea , therefore showing a selection for those conditions as postulated by several previous studies in salmon [25–27] . To investigate this we identified genotypes using the SNP assays in other year classes: 1999 for Eidselva ( 20 1SW , 8 3SW ) and 2004 for Suldalslågen ( 13 1SW , 13 3SW ) . Allele frequencies derived from the 11 SNP assays showed correlation to the 1SW and 3SW trait also in these year classes ( S3 Fig ) . Haplotype association analysis of these year classes again revealed two significant haplotypes also found in the 2008 year class ( Fig 3B ) . In the 1999 and 2004 year classes 44% ( β-value -0 . 96 , p-value = 3 . 77e-08 ) and 22% ( β-value 0 . 66 , p-value = 2 . 85e-04 ) of the phenotypic variation for age at maturity was explained by the 1SW or the 3SW haplotype , respectively . We thus conclude that genotypes at a single locus strongly influence sea age at maturity independent of year class across multiple salmon populations in Norway . Sea age at maturity can be significantly altered in salmon by modulating both light and temperature [1 , 28 , 29] . As a consequence , current aquaculture production methods include the use of constant light during the winter months to inhibit or reduce the incidence of early sexual maturation . The use of photoperiod to inhibit maturation in Norwegian farming has thereby masked the impact of this trait in commercial production . We were also interested to see how much the identified genetic trait contributed to the sea age at maturity trait in domesticated farmed salmon males , since wild salmon live in a different environment including different feed availability and water temperature that may trigger time of male puberty differently . To assay the linkage between phenotype and genotype for sea age at maturity in a domesticated strain , we utilized DNA from sexually maturing salmon from four different families of the Mowi strain . This strain has been in aquaculture for at least ten generations and has been selected for a variety of traits including growth and late maturation [30–32] . Mowi was originally obtained from a range of large wild salmon populations from Western Norway in 1969 , and has later been bred using a four-year life cycle . The long life cycle breeding has thereby probably increased the allele frequency for the late maturity phenotype . In this common garden experiment using the Mowi strain , fish were grown under natural light conditions in sea cages where males were matured after 1 , 2 or 3 or more years in sea . Haplotype analysis of these fish ( n = 97 ) revealed a shorter haplotype , consisting only of four SNPs , covering only 2386 bp in the 5’ end of the region assayed ( Fig 3C and S4 Fig ) . The observed differences between wild and domesticated fish may be due to the domestication process in this strain . These data clearly demonstrate that age at puberty can be explained by SNPs in this region also in a domesticated strain in culture for more than ten generations . Altogether the experiments clearly show that the selective sweep on Chr 25 significantly contributes to sea age at maturity both in wild and domesticated male salmon . Gene prediction in this area revealed three genes; charged multivesicular protein 2B ( chmp2B ) , vestigial-like protein 3 ( vgll3 ) and a-kinase anchor protein 11 ( akap11 , Fig 2B ) . From the analysis of domesticated salmon we could decrease the area of selection to a 2 . 4 kb region covering only vgll3 . For this locus we could identify loci containing paralogous genes . Two such loci were found tandemly repeated in Chr 21 . To assay if these two identical regions had SNPs associated with sea age at maturity we had to manually inspect both regions , using ambiguously mapped reads . No SNPs associated with the trait were discovered within these two paralogous regions . From our genotyping assay on domesticated samples we could with certainty reduce the region under selection in the downstream region of the vgll3 locus since we had genotyped several SNPs in this area . We can however not exclude that the upstream region of the vgll3 locus contained SNPs contributing to the haplotype since this area was not represented in our genotyping assay due to a large gap in the genome at this region ( Fig 2B ) . The 2 . 4 kb region contained two missense mutations in vgll3; at amino acid ( aa ) 54 and 323 . The sea age at maturity trait was strongly associated to the genotype of these SNPs since 36% ( nt . 28656101 Chr25 , β-value -0 . 61 , p-value = 9 . 80e-07 ) and 33% ( nt . 28658151 Chr 25 , β-value -0 . 60 , p-value = 3 . 77e-08 ) of the phenotypic variation could be explained by the genotype . The haplotypes associated with the 3SW trait encode a Thr and a Lys at these positions whereas the haplotype associated with 1SW encodes a Met and an Asp . Our analysis could not conclude whether these missense mutations are causative for the sea age at maturity trait , but since they occurred consistently together in the material we cannot rule out whether both or other non-coding variants at this locus are involved in age at maturity phenotype . It is also known from other studies that co-occurring amino acid changes can confer a phenotype [33] . The Vgll3 protein functions as a cofactor for the TEA Domain ( TEAD ) family of transcription factors [34] . The transcription factor binding region spanning aa105-aa134 in Vgll3 does not include any of the aa changes discovered which suggests that any direct binding differences between 1SW or 3SW fish are unlikely . It is thus difficult to predict how these amino acid changes affect the protein . At this point we cannot elucidate whether it is these missense mutations or other SNPs outside coding regions , which confer the trait variation . The question about the ancestral and derived alleles remains elusive , but we surveyed sequences from other salmonids for information about the amino acid variants and found that both brown trout ( Salmo trutta L . ) , rainbow trout ( Oncorhynchus mykiss ) and arctic char ( Salvelinus alpinus ) all have the 3SW variants of the amino acids . In addition , we ran an allelic discrimination assay on five individuals from the Swedish landlocked Atlantic salmon population , Gullspång ( landlocked for 10 , 000 years ) , all carrying only the 3SW ( Thr-Lys ) amino acid variant . This indicates that the 3SW version of the Vgll3 is ancestral and that the 1SW ( Met-Asp ) is derived . In humans the VGLL3 locus has been linked to age at maturity or puberty by a SNP in close proximity of the gene [35] , strengthening our notion that the salmon Vgll3 protein is involved in age at puberty in fish . Regarding the function of this protein in controlling age at maturity , it is known that Vgll3 is involved in the inhibition of adipocyte differentiation in mouse ( Mus musculus ) [36] . Changes in fat metabolism may be partially causative for changes in age at maturity , since increased adiposity has previously been linked to maturation in salmon [37–40] . In studies in rodent testis , vgll3 transcripts have been associated with differential expression during the early stages of steroidogenesis in the embryonic testis [41] , suggesting a role in testis maturation . Further functional studies of this protein and adjacent regulatory regions will confirm if the previous study in humans and our study have actually revealed a universal regulator of age at maturity in vertebrates . The most significant SNPs were located in the vgll3 locus but two neighboring genes , chmp2B and akap11 , also contain several significant SNPs ( Fig 2B ) . One of these , a missense mutation in akap11 translates to a Val in 1SW and a Met in 3SW at aa 214 . AKAP11 is involved in compartmentalization of cyclic AMP-dependent protein kinase ( PKA ) . This aa AKAP11 is not located in any of the known functional domains related to PKA [42] . AKAP 11 is highly expressed in elongating spermatocytes and mature sperm in human testis and is believed to contribute to cell cycle control in both germ cells and somatic cells . There are no reports clearly linking this protein to age at maturity but future functional studies may reveal if this is the case . Chmp2B did not contain any missense mutations but upon manual review of this region we detected a misplaced 16 , 885 bp region in the Chr 25 containing exon 1 and 2 of chmp2B . This region also carried many significant SNPs which were probably associated with the selective sweep . When this region was placed in proximity of the gene ( dark grey box in Fig 2B ) it became clear that many significant SNPs were near the chmp2b gene . This gene encodes a protein belonging to a protein complex which is involved in protein endocytosis [43] ( Fig 2B ) . In humans CHMP2B is known to be essential for the survival of nerve cells and is linked to both dementia and Amyotrophic lateral sclerosis ( ALS ) [44–46] . It is well known that the neural system works as a gatekeeper in controlling age of puberty , also in fish [47] but whether Chmp2B is involved in the regulation of puberty remains to be elucidated . In this study we performed a GWAS by genome re-sequencing with the aim to screen the genome of Atlantic salmon for loci regulating age at maturity in males . By investigating late and early maturing male fish from six rivers in Western Norway we demonstrated that the sea age at maturity trait was strongly associated with sequence variation at one locus on Chr 25 . The haplotype associated with late maturity can be used for selective breeding on individuals predisposed for this trait , thereby possibly reducing the incidence of negative phenotypes associated with early maturation of males in salmon aquaculture . However , using only late maturing fish in aquaculture breeding will increase generation times in culture , thereby decreasing the speed at which other traits such as growth can be selected for . This potential conflict of interest could be solved by using photoperiod manipulation to shorten generation time of fish with a genetic predisposition for high age at maturity . In this study we did not investigate how female maturation is affected by vgll3 genotypes . Future studies will reveal if female puberty is also influenced by this locus . This study also shows that certain haplotypes significantly contribute to the sea age at puberty , and may therefore be implemented as markers in the management of wild salmon populations in the face of changing environmental conditions such as increased sea temperatures . Significantly , this study and a previous study in humans [35] , suggests a conserved role of the Vgll3 protein in timing of puberty in vertebrates .
The samples of wild salmon upon which this study is based were collected by Rådgivende Biologer AS , Bergen , Norway ( http://www . radgivende-biologer . no ) . Scales were taken from dead salmon fish that had been captured by anglers during the fishing season . In this manner , samples of wild salmon were acquired from six rivers in Western Norway; Eidselva , Gloppenelven , Flekkeelven , Årdalselva , Suldalslågen and Vormo ( Fig 1 ) . In order to minimize the potential influence of environmental variation on the sea age at maturity we used fish from the same smolt year class sampled as 1SW fish ( returning to river 2009 ) and 3SW fish ( returning to river 2011 ) . Each river was represented by 20 1SW and 20 3SW males . For the genotyping assay we also included two other year classes from Eidselva and Suldalslågen . From Eidselva we retrieved scales from 20 1SW males from year 2000 and 8 3SW males from 2002 . From Suldalslågen we obtained scales from 14 1SW and 14 3SW from years 2005 and 2007 , respectively . In addition to samples of wild salmon , we investigated age at maturity in four full sibling families of domesticated salmon from the Norwegian Mowi strain maturing at 1SW , 2SW or older . These fish were obtained from an ongoing study at the Matre Aquaculture Research station where they were reared in a common garden design in sea cages without the use of continuous light , i . e . under ambient light only . Before transfer to sea cages , fish were sedated ( 0 . 07 gL-1 , Finquel , ScanAqua ) , adipose fin clipped and PIT ( passive integrated transponder ) tagged . Fin clips , preserved on 95% ethanol , from a total of 97 fish maturing at different sea ages were included in this study . The four families consisted of 36 , 24 , 13 and 24 sibling fish per family . We used the parental information to avoid within family Mendelian errors and to phase decendent’s haplotypes . Total DNA from selected individuals was purified from 2 to 3 scales using Qiagen DNeasy Blood & Tissue Kit ( Qiagen , Hilden , Germany ) according to the manufacturer´s recommendations . Sex of all samples used herein was validated by a PCR-based methodology aimed to detect the presence of the sdY gene [48 , 49] . Individuals showing amplicons of exon 2 and 4 were designated as males . As a positive PCR control and for species determination we used the presence of the 5S rRNA gene [50] . PCR amplifications were performed using reaction mixtures containing approximately 50 ng of extracted Atlantic salmon DNA , 10 nM Tris–HCl pH 8 . 8 , 1 . 5 mM MgCl2 , 50 mM KCl , 0 . 1% Triton X-100 , 0 . 35 μM of each primers , 0 . 5 Units of DNA Taq Polymerase ( Promega , Madison , WI , USA ) and 250 μM of each dNTP in a final volume of 20 μL . PCR products were visualized in 3% agarose gels . Following fluorometric quantification , equal amounts of DNA from ten males were pooled to generate paired-end libraries using the Genomic DNA Sample Preparation Kit ( Illumina , CA , USA ) according to manufacturer’s instructions . Libraries were sequenced on the Illumina HiSeq2000 platform ( Illumina , CA , USA ) at the Norwegian Sequencing center ( https://www . sequencing . uio . no/ , Oslo , Norway ) . In each sequencing lane we used pools of 10 fish from each sea age and river which made a total of 24 lanes sequenced in the whole experiment ( 6 rivers , 2 replicates per sea age ) . Raw sequence data has been deposited at SRA with BioProject Accession number PRJNA293012 . Library quality control was conducted to ensure that all the samples fulfilled the quality standards ( FastQC—http://www . bioinformatics . babraham . ac . uk/projects/fastqc/ ) . Adapter and quality trimming of FastQ format reads were carried out using Cutadapt [51] . All 24 libraries containing on average 361821757 ( ± 4956053 ) paired end reads were approved for further analysis and aligned to the most recent salmon genome release ( Acc . No . AGKD0000000 . 4 ) using Bowtie2 ( v . 2 . 1 . 0 ) [52] . Entire read alignment with no soft clipping was required by setting Bowtie 2 to the end-to-end mode . Seed length during alignment was set to 18 , allowing only 1 mismatch . Interval function between seed substring during multiseed alignment was defined by the following variables: S , 1 , 1 . 5 controlling the sensitivity of the mapping ( interval function f ( x ) = 1+1 . 5*sqrt ( x ) , x being the length of the read ) . Maximum number of ambiguous characters was set by the following function parameters: L , 0 , 0 . 1 ( f ( x ) = 0+0 . 1*x , x being read length ) . Minimum alignment score was governed by the function parameters L , -0 . 6 , -0 . 4 . ( f ( x ) = -0 . 6 + -0 . 4*x , x being read length ) . Only unambiguously mapped reads ( mapping quality score greater than 20 ) were retained for downstream analysis To improve the sensitivity to detect rare alleles , biological replicates in the dataset were bioinformatically fused using SAMtools merge , producing a single BAM file for each river and maturation stage [53] . SNPs were called using the Mpileup command in SAMtools . The resulting file was then recoded for use in the PoPoolation2 pipeline ( v . 1 . 2 . 2 ) [54] . A minimum base quality threshold of 20 was established in order to remove ambiguously mapped reads and low quality bases . The Cochran-Mantel-Haenszel test for repeated tests of independence for every SNP was performed using the PoPoolation 2 package ( cmh-test . pl and R based custom script ) in order to detect significant differences ( 0 . 1% FDR ) in allele frequencies between 1SW and 3SW pools [55] . For each merged sample of 20 fish , the parameters min-count was set to a value of 10 whereas the min-coverage and max-coverage were set to 7 and 42 , respectively ( 5–95% percentile ) . To annotate the salmon genome ( AKGD00000000 . 4 ) , Augustus gene prediction software was trained using PASA gene candidates by mapping salmon ESTs from NCBI to the salmon genome assembly with PASA [56 , 57] . The Augustus de novo gene prediction contained coding sequences without UTRs . The genes were validated by RNASeq from both Atlantic salmon [58] and rainbow trout [59] and annotated with Swissprot . Significant SNPs were functionally annotated to predict variant effect ( custom R and Python scripts ) . Bioinformatical analysis identified a 16 , 885 bp region in Chr 25 ( position 28907421–28924305 ) which contained the first two exons of the gene chmp2b in addition to several significant SNPs . In a previous version of the genome assembly ( AKGD00000000 . 3 ) this region existed as a single contig , and has presumably been inserted into the wrong chromosomal region in the most recent genome assembly . We corrected this by reverse-complementing the region and inserting it in the gap between the third exon of chmp2b and vgll3 in Chr 25 ( position 28626249–28643134 ) placing the exons of chmp2b in coherent order . Genotype/phenotype association analysis was performed using Plink v1 . 8 [24] . Selected SNPs were tested for association using a standard linear regression of phenotype on allele dosage ( Wald test 1%FDR [56] ) . Whenever possible , asymptotic haplotype-specific association tests were performed in order to establish the percentage of the phenotypic variation that could be explained by the detected haplotypes . The salmon genome has undergone a recent whole genome duplication and is partly tetraploid giving rise to many highly similar duplicated regions [20] . We have checked that all genotyping assays used in this study only targeted unique sequences in the genome . Eleven of the most significant SNPs ( S1 Table ) identified in the putative selective sweep ( Fig 2B ) were used to design a Sequenom assay . This was performed to be able to verify the GWAS findings for the 240 individuals used for pooled sequencing . In addition , we validated the associated SNPs in year classes 1999 and 2004 belonging to Eidselva and Suldalslågen rivers respectively as well as in 97 fish belonging to the 4 families from the domesticated Mowi strain . Genotyping was conducted on a Sequenom MassARRAY analyser ( San Diego , CA , USA ) . A complete list of Primers and extension primers used are found in the S2 Table . Sequenced pool material submitted to SRA ( Bioproject number PRJNA293012 ) . SNP data obtained has been deposited at http://marineseq . imr . no/salmatsnp/ . Scale samples from wild salmon were collected by local anglers during the fishing season , thus no permits/licenses regarding the collection of these samples were required by the research team . Samples from domesticated fish were retrieved from an ongoing study at Matre Aquaculture Research station ( IMR ) , where the experimental protocol ( permit number 4268 ) had been approved by the Norwegian Animal Research Authority ( NARA ) . Welfare and use of these experimental animals was performed in strict accordance with the Norwegian Animal Welfare Act of 19th of June 2009 , in force from 1st of January 2010 . All personnel involved in the experiment had undergone training approved by the Norwegian Food Safety Authority . This training is mandatory for all personnel running experiments involving animals included in the Animal Welfare Act . | For most species the factors that contribute to the genetic predisposition for age at maturity are currently unknown . In salmon aquaculture early maturation is negative for the growth , disease resistance and flesh quality . In addition , using populations of salmon selected to mature late may limit the genetic impact of aquaculture escapees , as these late maturing fish are more likely to die before they reach maturity . The aim of this study was to elucidate the genetic predisposition for salmon maturation . We determined the sequences of genomes from Atlantic salmon maturing early and late in six Norwegian rivers . This methodology enabled us to identify a short genomic region involved in determining the age at maturity in male Atlantic salmon . This region has also previously been linked to time of puberty in humans–supporting a general mechanism behind age at maturity in vertebrates . The results of this study may be used to breed salmon that are genetically predisposed to mature late which will improve welfare and production in aquaculture industry and aid in the management of escaped farmed salmon . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | The vgll3 Locus Controls Age at Maturity in Wild and Domesticated Atlantic Salmon (Salmo salar L.) Males |
Biologically inspired deep convolutional neural networks ( CNNs ) , trained for computer vision tasks , have been found to predict cortical responses with remarkable accuracy . However , the internal operations of these models remain poorly understood , and the factors that account for their success are unknown . Here we develop a set of techniques for using CNNs to gain insights into the computational mechanisms underlying cortical responses . We focused on responses in the occipital place area ( OPA ) , a scene-selective region of dorsal occipitoparietal cortex . In a previous study , we showed that fMRI activation patterns in the OPA contain information about the navigational affordances of scenes; that is , information about where one can and cannot move within the immediate environment . We hypothesized that this affordance information could be extracted using a set of purely feedforward computations . To test this idea , we examined a deep CNN with a feedforward architecture that had been previously trained for scene classification . We found that responses in the CNN to scene images were highly predictive of fMRI responses in the OPA . Moreover the CNN accounted for the portion of OPA variance relating to the navigational affordances of scenes . The CNN could thus serve as an image-computable candidate model of affordance-related responses in the OPA . We then ran a series of in silico experiments on this model to gain insights into its internal operations . These analyses showed that the computation of affordance-related features relied heavily on visual information at high-spatial frequencies and cardinal orientations , both of which have previously been identified as low-level stimulus preferences of scene-selective visual cortex . These computations also exhibited a strong preference for information in the lower visual field , which is consistent with known retinotopic biases in the OPA . Visualizations of feature selectivity within the CNN suggested that affordance-based responses encoded features that define the layout of the spatial environment , such as boundary-defining junctions and large extended surfaces . Together , these results map the sensory functions of the OPA onto a fully quantitative model that provides insights into its visual computations . More broadly , they advance integrative techniques for understanding visual cortex across multiple level of analysis: from the identification of cortical sensory functions to the modeling of their underlying algorithms .
Recent advances in the use of deep neural networks for computer vision have yielded image computable models that exhibit human-level performance on scene- and object-classification tasks [1–4] . The units in these networks often exhibit response profiles that are predictive of neural activity in mammalian visual cortex [5–11] , suggesting that they might be profitably used to investigate the computational algorithms that underlie biological vision [12–16] . However , many of the internal operations of these models remain mysterious , and the fundamental theoretical principles that account for their predictive accuracy are not well understood [16–18] . This presents an important challenge to the field: if deep neural networks are to fulfill their potential as a method for investigating visual perception in living organisms , it will first be necessary to develop techniques for using these networks to provide computational insights into neurobiological systems . It is this issue—the use of deep neural networks for gaining insights into the computational processes of biological vision—that we address here . We focus in particular on the mechanisms underlying natural scene perception . A central aspect of scene perception is the identification of the navigational affordances of the local environment—where one can move to ( e . g . , a doorway or an unobstructed path ) , and where one's movement is blocked . In a recent fMRI study , we showed that the navigational-affordance structure of scenes could be decoded from multivoxel response patterns in scene-selective visual areas [19] . The strongest results were found in a region of the dorsal occipital lobe known as the occipital place area ( OPA ) , which is one of three patches of high-level visual cortex that respond strongly and preferentially to images of spatial scenes [20–24] . These results demonstrated that the OPA encodes affordance-related visual features . However , they did not address the crucial question of how these features might be computed from sensory inputs . There was one aspect of the previous study that provided a clue as to how affordance representations might be constructed: affordance information was present in the OPA even though participants performed tasks that made no explicit reference to this information . For example , in one experiment , participants were simply asked to report the colors of dots overlaid on the scene , and in another experiment , they were asked to perform a category-recognition task . Despite the fact that these tasks did not require the participants to think about the spatial layout of the scene or plan a route through it , it was possible to decode navigational affordances in the OPA in both cases . This suggested to us that affordances might be rapidly and automatically extracted through a set of purely feedforward computations . In the current study we tested this idea by examining a biologically inspired CNN with a feedforward architecture that was previously trained for scene classification [3] . This CNN implements a hierarchy of linear-nonlinear operations that give rise to increasingly complex feature representations , and previous work has shown that its internal representations can be used to predict neural responses to natural scene images [25 , 26] . It has also been shown that the higher layers of this CNN can be used to decode the coarse spatial properties of scenes , such as their overall size [25] . By examining this CNN , we aimed to demonstrate that affordance information could be extracted by a feedforward system , and to better understand how this information might be computed . To preview our results , we find that the CNN contains information about fine-grained spatial features that could be used to map out the navigational pathways within a scene; moreover , these features are highly predictive of affordance-related fMRI responses in the OPA . These findings demonstrate that the CNN can serve as a candidate , image-computable model of navigational-affordance coding in the human visual system . Using this quantitative model , we then develop a set of techniques that provide insights into the computational operations that give rise to affordance-related representations . These analyses reveal a set of stimulus input features that are critical for predicting affordance-related cortical responses , and they suggest a set of high-level , complex features that may serve as a basis set for the population coding of navigational affordances . By combining neuroimaging findings with a fully quantitative computational model , we were able to complement a theory of cortical representation with discoveries of its algorithmic implementation—thus providing insights at multiple levels of understanding and moving us toward a more comprehensive functional description of visual cortex .
To test for the representation of navigational affordances in the human visual system , we examined fMRI responses to 50 images of indoor environments with clear navigational paths passing through the bottom of the scene ( Fig 1A ) . Subjects viewed these images one at a time for 1 . 5 s each while maintaining central fixation and performing a category-recognition task that was unrelated to navigation ( i . e . , press a button when the viewed scene was a bathroom ) . Details of the experimental paradigm and a complete analysis of the fMRI responses can be found in a previous report [19] . In this section , we briefly recapitulate the aspects of the results that are most relevant to the subsequent computational analyses . To measure the navigational affordances of these stimuli , we asked an independent group of subjects to indicate with a computer mouse the paths that they would take to walk through each environment starting from the bottom of the image ( Fig 1B ) . From these responses , we created probabilistic maps of the navigational paths through each scene . We then constructed histograms of these navigational probability measurements in one-degree angular bins over a range of directions radiating from the starting point of the paths . These histograms approximate a probabilistic affordance map of potential navigational paths radiating from the perspective of the viewer [27] . We then tested for the presence of affordance-related information in fMRI responses using representational similarity analysis ( RSA ) [28] . In RSA , the information encoded in brain responses is compared with a cognitive or computational model through correlations of their representational dissimilarity matrices ( RDMs ) . RDMs are constructed through pairwise comparisons of the model representations or brain responses for all stimulus classes ( in this case , the 50 images ) , and they serve as a summary measurement of the stimulus-class distinctions . The correlation between any two RDMs reflects the degree to which they contain similar information about the stimuli . We constructed an RDM for the navigational-affordance model through pairwise comparisons of the affordance histograms ( Fig 1C ) . Neural RDMs were constructed for several regions of interest ( ROIs ) through pairwise comparisons of their multivoxel activation patterns for each image . We focused our initial analyses on three ROIs that are known to be strongly involved in scene processing: the OPA , the parahippocampal place area ( PPA ) , and the retrosplenial complex ( RSC ) [20–24] . All three of these regions respond more strongly to spatial scenes ( e . g . , images of landscapes , city streets , or rooms ) than other visual stimuli , such as objects and faces , and thus are good candidates for supporting representations of navigational affordances . We also examined patterns in early visual cortex ( EVC ) . Using RSA to compare the RDMs for these regions to the navigational-affordance RDM , we found evidence that affordance information is encoded in scene-selective visual cortex , most strongly in the dorsal scene-selective region known as the OPA ( Fig 1C ) . These effects were not observed in lower-level EVC , suggesting that navigational affordances likely reflect mid-to-high-level visual features that require several computational stages along the cortical hierarchy . In our previous report , a whole-brain searchlight analysis confirmed that the strongest cortical locus of affordance coding overlapped with the OPA [19] . Interestingly , affordance coding in scene regions was observed even though participants performed a perceptual-semantic recognition task in which they were not explicitly asked about the navigational affordances of the scene—suggesting that affordance information is automatically elicited during scene perception . Together , these results suggest that scene-selective visual cortex routinely encodes complex spatial features that can be used to map out the navigational affordances of the local visual scene . These analyses provide functional insights into visual cortex at the level of representation—that is , the identification of sensory information encoded in cortical responses . However , an equally important question for any theory of sensory cortical function is to understand how its representations can be computed at an algorithmic level [12–16] . Understanding the algorithms that give rise to high-level sensory representations requires a quantitative model that implements representational transformations from visual stimuli . Thus , we next turn to the question of how affordance representations might be computed from sensory inputs . Visual cortex implements a complex set of highly nonlinear transformations that remain poorly understood . Attempts at modeling these transformations using hand-engineered algorithms have long fallen short of accurately predicting mid-to-high-level sensory representations [6 , 10 , 11 , 29–31] . However , advances in the development of artificial deep neural networks have dramatically changed the outlook for the quantitative modeling of visual cortex . In particular , recently developed deep CNNs for tasks such as image classification have been found to predict sensory responses throughout much of visual cortex at an unprecedented level of accuracy [5–11] . The performance of these CNNs suggests that they hold the promise of providing fundamental insights into the computational algorithms of biological vision . However , because their internal representations were not hand-engineered to test specific theoretical operations , they are challenging to interpret . Indeed , most of the critical parameters in CNNs are set through supervised learning for the purpose of achieving accurate performance on computer vision tasks , meaning that the resulting features are unconstrained by a priori theoretical principles . Furthermore , the complex transformations of these internal CNN units cannot be understood through a simple inspection of their learned parameters . Thus , neural network models have the potential to be highly informative to sensory neuroscience , but a critical challenge for moving forward is the development of techniques to probe the factors that best account for similarities between cortical responses and the internal representations of the models . Here we tested a deep CNN as a potential candidate model of affordance-related responses in scene-selective visual cortex . Given the apparent automaticity of affordance-related responses , we hypothesized that they could be modeled through a set of purely feedforward computations performed on image inputs . To test this idea , we examined a model that was previously trained to classify images into a set of scene categories [3] . This feedforward model contains 5 convolutional layers followed by 3 fully connected layers , the last of which contains units corresponding to a set of scene category labels ( Fig 2A ) . The architecture of the model is similar to the AlexNet model that initiated the recent surge of interest in CNNs for computer vision [2] . Units in the convolutional layers of this model have local connectivity , giving rise to increasingly large spatial receptive fields from layers 1 through 5 . The dense connectivity of the final three layers means that the selectivity of their units could depend on any spatial position in the image . Each unit in the CNN implements a linear-nonlinear operation in which it computes a weighted linear sum of its inputs followed by a nonlinear activation function ( specifically , a rectified linear threshold ) . The weights on the inputs for each unit define a type of filter , and each convolutional layer contains a set of filters that are replicated with the same set of weights over all parts of the image ( hence the term “convolution” ) . There are two other nonlinear operations implemented by a subset of the convolutional layers: max-pooling , in which only the maximum activation in a local pool of units is passed to the next layer , and normalization , in which activations are adjusted through division by a factor that reflects the summed activity of multiple units at the same spatial position . Together , this small set of functional operations along with a set of architectural constraints define an untrained model whose many other parameters can be set through gradient descent with backpropagation—producing a trained model that performs highly complex feats of visual classification . We passed the images from the fMRI experiment through the CNN and constructed a set of RDMs using the final outputs from each layer . We then used RSA to compare the representations of the CNN with: ( i ) the RDM for the navigational-affordance model and ( ii ) the RDM for fMRI responses in the OPA . The RSA comparisons with the affordance model showed that the CNN contained affordance-related information , which arose gradually across the lower layers and peaked in layer 5 , the highest convolutional layer ( Fig 2B ) . Note that this was the case despite the fact that the CNN was trained to classify scenes based on their categorical identity ( e . g . , kitchen ) , not their affordance structure . Weak effects were observed in lower convolutional layers , consistent with the pattern of findings from the fMRI experiment , in which affordance representations were not evident in EVC , and they suggest that affordances reflect mid-to-high-level , rather than low-level , visual features . The decrease in affordance-related information in the last three fully connected layers may result from the increasingly semantic nature of representations in these layers , which ultimately encode a set of scene-category labels that are likely unrelated to the affordance-related features of the scenes . The RSA comparisons with OPA responses showed that the CNN provided a highly accurate model of representations in this brain region , with strong effects across all CNN layers and a peak correlation in layer 5 ( Fig 2B ) . Indeed , several layers of the CNN reached the highest accuracy we could expect for any model , given the noise ceiling of the OPA , which was calculated from the variance across subjects ( r-value for OPA noise ceiling = 0 . 30 ) . Together , these findings demonstrate the feasibility of computing complex affordance-related features through a set of purely feedforward transformations , and they show that the CNN is a highly predictive model of OPA responses to natural images depicting such affordances . The above findings demonstrate that the CNN is representationally similar to the navigational-affordance RDM and also similar to the OPA RDM , but they leave open the important question of whether the CNN captures the same variance in the OPA as the navigational-affordance RDM . In other words , can the CNN serve as a computational model for affordance-related responses in the OPA ? To address this question , we combined the RSA approach with commonality analysis [32] , a variance partitioning technique in which the explained variance of a multiple regression model is divided into the unique and shared variance contributed by all of its predictors . In this case , multiple regression RSA was used to construct an encoding model of OPA representations . Thus , the OPA was the predictand and the affordance and CNN models were predictors . Our goal was to identify the portion of the shared variance between the affordance RDM and OPA RDM that could be accounted for by the CNN RDM ( Fig 3A ) . This analysis showed that the CNN could explain a substantial portion of the representational similarity between the navigational-affordance model and the OPA . In particular , over half of the explained variance of the navigational-affordance RDM could be accounted for by layer 5 of the CNN ( Fig 3B ) . This suggests that the CNN can serve as a candidate , quantitative model of affordance-related responses in the OPA . One of the most important aspects of the CNN as a candidate model of affordance-related cortical responses is that it is image computable , meaning that its representations can be calculated for any input image . This makes it possible to test predictions about the internal computations of the model by generating new stimuli and running in silico experiments . In the next two sections , we run a series of experiments on the CNN to gain insights into the factors that underlie its predictive accuracy in explaining the representations of the navigational-affordance model and the OPA . A fundamental issue for understanding any model of sensory computation is determining the aspects of the sensory stimulus on which it operates . In other words , what sensory inputs drive the responses of the model ? To answer this question , we investigated the image features that drive affordance-related responses in the CNN . Specifically , we sought to identify classes of low-level stimulus features that are critical for explaining the representational similarity of the CNN to the navigational-affordance model and the OPA . We expected that navigational affordances would rely on image features that convey information about the spatial structure of scenes . Our specific hypotheses were that affordance-related representations would be relatively unaffected by color information and would rely heavily on high spatial frequencies and edges at cardinal orientations ( i . e . , horizontal and vertical ) . The hypothesis that color information would be unimportant was motivated by our intuition that color is not typically a defining feature of the structural properties of scenes and by a previous finding of ours showing that affordance representations in the OPA are partially tolerant to variations in scene textures and colors [19] . The other two hypotheses were motivated by previous work suggesting that high spatial frequencies and cardinal orientations are especially informative for the perceptual analysis of spatial scenes , and that the PPA and possibly other scene-selective regions are particularly sensitive to these low-level visual features [33–38] , but see [39] . To test these hypotheses , we generated new sets of filtered stimuli in which specific visual features were isolated or removed ( i . e . , color , spatial frequencies , cardinal or oblique edges; Fig 4A and 4B ) . These filtered stimuli were passed through the CNN , and new RDMs were created for each layer . We used the commonality-analysis technique described in the previous section to quantify the portion of the original explained variance of the CNN that could be accounted for by the filtered stimuli . This procedure was applied to the explained variance of the CNN for predicting both the navigational-affordance RDM and the OPA RDM ( Fig 4A ) . The results for both sets of analyses showed that over half of the explained variance of the CNN could be accounted for when the inputs contained only grayscale information , high-spatial frequencies , or edges at cardinal orientations . In contrast , when input images containing only low-spatial frequencies or oblique edges were used , a much smaller portion of the explained variance was accounted for . The differences in explained variance across high and low spatial frequencies and across cardinal and oblique orientations were more pronounced for the RSA predictions of the affordance RDM , but a similar pattern was observed for the OPA RDM . We used a bootstrap resampling procedure to statistically assess these comparisons . Specifically , we calculated bootstrap 95% confidence intervals for the following contrasts of shared-variance scores: 1 ) high spatial frequencies minus low spatial frequencies and 2 ) cardinal orientations minus oblique orientations . These analyses showed that the differences in shared variance for high vs . low spatial frequencies and for cardinal vs . oblique orientations were reliable for both the affordance RDM and the OPA RDM ( all p<0 . 05 , bootstrap ) . Together , these results suggest that visual inputs at high-spatial frequencies and cardinal orientations are important for computing the affordance-related features of the CNN . Furthermore , these computational operations appear to be largely tolerant to the removal of color information . Indeed , it is striking how much explained variance these inputs account for given how much information has been discarded from their corresponding filtered stimulus sets . In addition to examining classes of input features to the CNN , we also sought to understand how inputs from different spatial positions in the image affected the similarity between the CNN and RDMs for the navigational-affordance model and the OPA . Our hypothesis was that these RSA effects would be driven most strongly by inputs from the lower visual field ( we use the term “visual field” here because the fMRI subjects were asked to maintain central fixation throughout the experiment ) . This was motivated by previous findings showing that the OPA has a retinotopic bias for the lower visual field [40 , 41] and the intuitive prediction that the navigational affordances of local space rely heavily on features close to the ground plane . To test this hypothesis , we generated sets of occluded stimuli in which everything except a small horizontal slice of the image was masked ( Fig 5 ) . These occluded stimuli were passed through the CNN , and new RDMs were created for each layer . Once again , we used the commonality-analysis technique described above to quantify the portion of the original explained variance of the CNN that could still be accounted for by these occluded stimuli . This procedure was repeated with the un-occluded region slightly shifted on each iteration until the entire vertical extent of the image was sampled . We used this procedure to analyze the explained variance of the CNN for predicting both the navigational-affordance RDM and the OPA RDM ( Fig 5 ) . For comparison , we also applied this procedure to RDMs for the other ROIs . These analyses showed that the predictive accuracy of the CNN for both the affordance model and the OPA was driven most strongly by inputs from the lower visual field . Strikingly , as much as 70% of the explained variance of the CNN in the OPA could be accounted for by a small horizontal band of features at the bottom of the image ( Fig 5 ) . We created a summary statistic for this visual-field bias by calculating the difference in mean shared variance across the lower and upper halves of the image . A comparison of this summary statistic across all tested RDMs shows that the lower visual field bias was observed for the RSA predictions of the affordance model and the OPA , but not for the other ROIs ( Fig 5 ) . Together , these results demonstrate that information from the lower visual field is critical to the performance of the CNN in predicting the affordance RDM and the OPA RDM . These findings are consistent with previous neuroimaging work on the retinotopic biases of the OPA [40 , 41] , and they suggest that the cortical computation of affordance-related features reflects a strong bias for inputs from the lower visual field . The analyses above examined the stimulus inputs that drive affordance-related computations in the CNN . We next set out to characterize the high-level features that result from these computations . Specifically , we sought to characterize the internal representations of the CNN that best account for the representations of the OPA and the navigational-affordance model . To do this , we performed a set of visualization analyses to reify the complex visual motifs detected by the internal units of the CNN . We characterized the feature selectivity of CNN units using a receptive-field mapping procedure ( Fig 6A ) [42] . The goal was to identify natural image features that drive the internal representations of the CNN . In this procedure , the selectivity of individual CNN units was mapped across each image by iteratively occluding the inputs to the CNN . First , the original , un-occluded image was passed through the CNN . Then a small portion of the image was occluded with a patch of random pixel values ( 11 pixels by 11 pixels ) . The occluded image was passed though the CNN , and the discrepancies in unit activations relative to the original image were logged . These discrepancy values were calculated for each unit by taking the difference in magnitude between the activation to the original image and the activation to the occluded image . After iteratively applying this procedure across all spatial positions in the image , a two-dimensional discrepancy map was generated for each unit and each image ( Fig 6A ) . Each discrepancy map indicates the sensitivity of a CNN unit to the visual information across all spatial positions of an image . The spatial distribution of the discrepancy effects reflects the position and extent of a unit’s receptive field , and the magnitude of the discrepancy effects reflects the sensitivity of a unit to the underlying image features . We focused our analyses on the units in layer 5 , which was the layer with the highest RSA correlation for the both the navigational-affordance model and the OPA . We selected 50 units in this layer based on their unit-wise RSA correlations to the navigational-affordance model and the OPA . These units were highly informative for our effects of interest: an RDM created from just these 50 units showed comparable RSA correlations to those observed when using all units in layer 5 ( correlation with affordance RDM: r = 0 . 28; correlation with OPA RDM: r = 0 . 35 ) . We generated receptive-field visualizations for each of these units . These visualizations were created by identifying the top 3 images that generated the largest discrepancy values in the receptive-field mapping procedure ( i . e . , images that were strongly representative of a unit’s preferences ) . A segmentation mask was then applied to each image by thresholding the unit’s discrepancy map at 10% of the peak discrepancy value . Segmentations highlight the portion of the image that the unit was sensitive to . Each segmentation is outlined in red , and regions of the image outside of the segmentation are darkened ( Fig 6B ) . We sought to identify prominent trends across this set of receptive-field segmentations . In a simple visual inspection of the segmentations , we detected visual motifs that were common among the units , and the results of an automated clustering procedure highlighted these trends . Using data-driven techniques , we embedded the segmentations into a low-dimensional space and then partitioned them into clusters with similar visual motifs . We used t-distributed stochastic neighbor embedding ( t-SNE ) to generate a two-dimensional embedding of the units based on the visual similarity of their receptive-field segmentations ( Fig 6B ) . We then used k-means clustering to identify sets of units with similar embeddings . The number of clusters was set at 7 based on the outcome of a cluster-evaluation procedure . The specific cluster assignments do not necessarily indicate major qualitative distinctions between units . Rather , they provide a data-driven means of reducing the complexity of the results and highlighting the broad themes in the data . These themes can also be seen in the complete set of visualizations plotted in S1–S7 Figs . These visualizations revealed two broad visual motifs: boundary-defining junctions and large , extended surfaces . Boundary-defining junctions are the regions of an image where two or more extended planes meet ( e . g . , clusters 1 , 5 , 6 , and 7 in Fig 6B ) . These were often the junctions of walls and floors , and less often ceilings . This was the most common visual motif across all segmentations . Large , extended surfaces were uninterrupted portions of floor and wall planes ( e . g . , cluster 3 in Fig 6B ) . There were also units that detected more complex structural features that were often indicative of doorways and other open pathways ( e . g . , clusters 2 and 4 in Fig 6B ) . A common thread running through all these visualizations is that they appear to reflect high-level scene features that could be reliably used to map out the spatial layout and navigational affordances of the local environment . Boundary-defining junctions and large , extended surfaces provide critical information about the spatial geometry of the local scene , and more fine-grained structural elements , such as doorways and open pathways , are critical to the navigational layout of a scene . Together , these results suggest a minimal set of high-level visual features that are critical for modeling the navigational affordances of natural images and predicting the affordance-related responses of scene-selective visual cortex . Our analyses thus far have focused on a carefully selected set of indoor scenes in which the potential for navigation was clearly delimited by the spatial layout of impassable boundaries and solid open ground . Indeed , the built environments depicted in our stimuli were designed so that humans could readily navigate through them . However , there are many environments in which navigability is determined by a more complex set of perceptual factors . For example , in outdoor scenes navigability can be strongly influenced by the material properties of the ground plane ( e . g . , grass , water ) . We wondered whether the components of the CNN that were related to the navigational-affordance properties of our indoor scenes could be used to identify navigational properties in a broader range of images . To address this question , we examined a set of images depicting natural landscapes , whose navigational properties had been quantified in a previous behavioral study [43] . Specifically , these stimuli included 100 images that could be grouped into categories of low or high navigability based on subjective behavioral assessments . The overall navigability of the images reflected subjects’ judgments of how easily they could move through the scene . These navigability assessments were influenced by a wide range of scene features , including the spatial layout of pathways and boundaries , the presence of clutter and obstacles , and the potential for treacherous conditions . There were also low and high categories for 13 other scene properties ( Fig 7A ) . Each scene property was associated with 100 images ( 50 low and 50 high ) , and many images were used for multiple scene properties ( 548 images in total ) . We sought to determine whether the units from the CNN that were highly informative for identifying the navigational affordances of indoor scenes could also discern the navigational properties in this heterogeneous set of natural landscapes . To do this , we focused on the 50 units selected for the visualization analyses in Fig 6 , and we used the responses of these units to classify natural landscapes based on their overall navigability ( Fig 7A ) . We found that not only were these CNN units able to classify navigability across a broad range of outdoor scenes , but they also appeared to be particularly informative for this task relative to the other units in layer 5 ( 99th percentile in a resampling distribution ) . Furthermore , these units were substantially better at classifying navigability than any other scene property ( chi-squared tests for equality of proportions: all p<0 . 05 , Bonferroni corrected ) . These findings suggest that the navigation-related components of the CNN detected in our previous analyses can generalize beyond the domain of built environments to identify the navigability of complex natural landscapes , whose navigational properties reflect a diverse set of high-level perceptual cues . Example images from the classification analysis of navigability are shown in Fig 7B and 7C . These images were classified based on the responses of the affordance-related CNN units . The correctly classified images span a broad range of semantic categories and have complex spatial and navigational properties ( Fig 7B ) . Spatial layout alone appears to be insufficient for explaining the performance of this classifier . For example , the high-navigability scenes could have either open or closed spatial geometries , and many of the low-navigability scenes have open ground but would be challenging to navigate because of the materials or obstacles that they contain ( e . g . , water , dense brush ) . This suggests that the affordance-related components of the CNN are sensitive to scene properties other than coarse-scale spatial layout and could potentially be used to detect a broader set of navigational cues , including textures and material properties . However , textures and materials also appeared to underlie some of the confusions of the classifier ( Fig 7C ) . For example , there were multiple instances of scenes covered in dense natural materials that the classifier labeled as low navigability , even though human observers considered them to be readily navigable ( e . g . , a field of flowers ) . These classification errors may reflect the limitations of the CNN units for making fine-grained distinctions between navigable and non-navigable materials . However , the materials and textures in these scenes were relatively uncommon among the stimuli , which suggests that a more important underlying factor may be the limited number of relevant training examples . Overall , these classification analyses demonstrate that the CNN contains internal representations that can be used to identify the navigational properties of both regularly structured built environments and highly-varied natural landscapes . These representations appear to be sensitive to a diverse set of high-level scene features , including not only spatial layout but also scene textures and materials .
These findings have important implications for developing a computational understanding of scene-selective visual cortex . ( Note that we use the term “computational” here in its conventional sense and not in the specific sense defined by David Marr [12] . For readers who are familiar with Marr’s levels of description , our analyses can be viewed as largely addressing the algorithmic level . ) To gain an understanding of the algorithms implemented by the visual system we first need candidate quantitative models whose parameters and operations can be interpreted for theoretical insights . One of the primary criteria for evaluating such a model is that it explains a substantial portion of stimulus-driven activity in the brain region of interest . Any model that does not meet this necessary criterion is fundamentally insufficient or incorrect . A major strength of the CNN examined here is that it is highly accurate at predicting cortical responses to the perception of natural scenes . Indeed , the CNN explained as much variance in the responses of the OPA as could be expected for any model , after accounting for the portion of OPA variance that could be attributed to noise . Thus , as in previous studies of high-level object perception , the ability of the CNN to reach the noise ceiling for explained variance during scene perception constitutes a major advance in the quantitative modeling of cortical responses [6 , 7] . Another strength of the CNN as a candidate model is that its representations can be computed from arbitrary image inputs . This image computability confers two major benefits . First , the internal representations of the CNN can be investigated across all computational stages and mapped onto a cortical hierarchy , allowing for a complete description of the nonlinear transformations that convert sensory inputs into high-level visual features [11] . Second , image computability allows investigators to submit novel stimulus inputs to the CNN for the purpose of testing hypotheses through in silico experiments . Here we took advantage of this image computability to test several hypotheses about which stimulus inputs are critical for computing affordance-related visual features and predicting the responses of the OPA . These analyses demonstrated the importance of inputs from the lower visual field ( i . e . , the bottom of the image when fixation is at the center ) , which aligns with previous fMRI studies that used receptive-field mapping to identify a lower-field bias in the OPA [40 , 41] . These analyses also demonstrated the importance of several low-level image features that have previously been shown to drive the responses of scene-selective visual cortex , including high-spatial frequencies and contours at cardinal orientations [33–38] , but see [39] . We also performed visualization experiments on the internal representations of the CNN to identify potential affordance-related scene features that might be encoded in the population responses of the OPA . Our approach involved data-driven visualizations of the image regions detected by individual CNN units . We focused on the fifth convolutional layer of the CNN and , in particular , on units in this layer that corresponded most strongly to the representations of the OPA and the navigational-affordance model . Among the scene features detected by these units , two broad themes were prominent: boundary-defining junctions and large , extended surfaces . Boundary-defining junctions were contours where two or more large and often orthogonal surfaces were adjoined . These included extended junctions of two surfaces , such as a wall and a floor , and corners where three surfaces come together , such as two walls and a floor . Thus , boundary-defining junctions resembled the basic features that one would use to sketch the spatial layout of a scene . The idea that such features might be encoded in scene-selective visual cortex accords with previous findings from neuroimaging and electrophysiology . The most directly related findings come from a series of neuroimaging studies investigating the responses of scene-selective cortex to line drawings of natural images [44 , 45] . These line-drawing stimuli convey information about contours and their spatial arrangement , but they lack many of the rich details of natural images , including color , texture , and shading . Nonetheless , these stimuli elicit representations of scene-category information in scene-selective visual cortex . These effects appear to be driven mostly by long contours and their junctions , whose arrangement conveys information about the spatial structure of a scene . This suggests that a substantial portion of the features encoded by scene-selective cortex can be computed using only structure-defining contour information . This aligns with the findings from our visualization analyses , which suggest that large surface junctions are an important component of the information encoded by scene-selective visual cortex . These surface junctions correspond to the long structure-defining contours that would be highlighted in a line drawing . Electrophysiological investigations of scene-selective visual cortex in the macaque brain have also demonstrated the importance of structure-defining contours [46] , and even identified cells that exhibited selectivity for the surface junctions in rooms , which appears to be remarkably similar to the selectivity for boundary-defining junctions identified here . Our findings are also broadly consistent with previous behavioral studies demonstrating that contours and contour junctions in 2D images are highly informative about the arrangement of surfaces in 3D space [47] . In particular , contour junctions convey information about 3D structure that is largely invariant to changes in viewpoint , making them exceptionally useful for inferring spatial structure [47] . In addition to boundary-defining junctions , we also observed selectivity for large extended surfaces . One recent study has suggested that the responses of scene-selective visual cortex can be well predicted from the depth and orientation of large surfaces in natural scenes [48] . This appears to be consistent with our finding , but other possible roles for the surface-preferring units in our study include texture identification and the use of texture gradients as 3D orientation cues [49 , 50] . Overall , this pattern of selectivity for large surfaces and the junctions between them is consistent with an information-processing mechanism for representing the spatial structure of the local visual environment [12] . However , the findings from our analysis of natural landscapes suggest that the affordance-related components of the CNN are additionally sensitive to other navigational cues , such as the textures and materials in outdoor scenes ( e . g . , water , grass ) . Thus , these affordance-related units may encode multiple aspects of the space-defining surfaces in a scene , including their structure and their material properties . These computational findings have implications for interpreting previous neuroimaging studies of scene-selective cortex . It has been argued that the apparent category selectivity of scene regions can be explained more parsimoniously in terms of preferences for low-level image features , such as high spatial frequencies [33–37] , but see [39] . However , the analyses presented here suggest an alternative interpretation , namely that scene-selective visual regions encode complex features that convey information about high-level scene properties , such as navigational layout and category membership , but that the computations that give rise to these features rely heavily on specific sets of low-level inputs [38 , 51] . This account characterizes the function of scene-selective visual cortex within the context of a computational system , and it demonstrates how a region within this system could exhibit response preferences for the low-level features that drive its upstream inputs . Thus , by examining a candidate computational model , we identified a potential mechanism through which neuroimaging studies could produce seemingly contradictory findings on the feature selectivity of scene-selective cortex . More broadly , these analyses demonstrate the importance of building explicit computational models to evaluate functional theories of high-level visual cortex . Doing so allows investigators to interpret cortical processes in terms of their functional significance to systems-level computations rather than region-specific representational models . The analyses and techniques presented here are broadly relevant to research on the functions of visual cortex . A major goal of visual neuroscience is to understand the information-processing mechanisms that visual cortex carries out [12] . Progress toward this goal can be assessed by how well investigators are able to implement these mechanisms de novo using models that reflect a compact set of theoretical principles . To this end , investigators require models that are constructed from mathematical algorithms , to allow for implementations in any suitable computational hardware , and whose internal operations are theoretically interpretable , in the sense that one can provide summary descriptions of the functions they carry out and the theoretical principles they embody . A long line of work in visual neuroscience has attempted to understand the information-processing mechanisms of visual cortex by hand-engineering computational models based on a priori theoretical principles [29 , 30 , 52] . Although this approach has been fruitful in characterizing the earliest stages of visual processing , it has not proved effective for explaining the functions of mid-to-high-level visual cortex , where the complexity of the operations and the number of possible features grows exponentially [11] . Recent advances in the development of deep CNNs trained for computer vision have incidentally yielded quantitative models that are remarkably accurate at predicting functional activity throughout much of the visual system [5–11] . However , from a theoretical perspective , these highly complex models have remained largely opaque , and little is known about what aspects of these models might be relevant for understanding the information processes of biological vision . Here we developed an approach for probing the internal operations of a CNN for insights into cortical computation . Our approach uses RSA in the context of multiple linear regression to evaluate similarities between computational , theoretical , and neural systems . A major benefit of RSA is that it evaluates the information in these systems through summary representations in RDMs , which avoids the many difficulties of identifying mappings between the individual units of high-dimensional systems [28] . We evaluated these multiple linear regressions using a variance-partitioning procedure that allowed us to quantify the degree to which representational models explained shared or unique components of the information content in a cortical region . These statistical methods were combined with techniques for running in silico experiments to test theoretically motivated hypotheses about information processing in the CNN and its relationship to the functions of visual cortex . Although we applied these techniques to an exploration of affordance coding in visual scenes , they are broadly applicable and could be used to examine any cortical region or image-computable model . They demonstrate a general approach for exploring how the computations of a CNN relate to the information-processing algorithms of biological vision . It is worth noting , however , that there are important limitations in using deep neural networks for insights into neurobiological processes . One of the critical limiting factors of the analyses described here is our reliance on a pre-trained computational model . Deep CNNs have large numbers of parameters , which are typically fit through supervised learning using millions of labeled stimuli . Given the cost of manually labeling this number of stimuli and the far smaller number of stimuli used in a typical neuroscience experiment , it is not feasible to train deep neural networks that are customized for the perceptual processes of interest in every new experiment . Fortunately , neuroscientists can take advantage of the fact that deep neural networks trained for real-world tasks using large , naturalistic stimulus sets appear to learn a set of general-purpose representations that often transfer well to other tasks [25 , 53 , 54] . Furthermore , the objective functions that these CNNs were trained for all relate to computer-vision goals ( e . g . , object or scene classification ) , and , yet , their internal representations exhibit remarkable similarities to those at multiple levels of visual cortex [5–11] . This means that investigators can examine existing pre-trained models for their potential relevance to a cortical sensory process , even if the models were not explicitly trained to implement that process . However , in the analysis of pre-trained models , the architecture , activation function , and other design factors are constrained , and , thus , the results of these analyses cannot be easily compared with alternative algorithmic implementations . An important direction for future work will be the use of multiple models to compare specific architectural and design factors with neural processes , such as the number of model layers , the directions and patterns of connectivity between neurons , the kinds of non-linear operations that the neurons implement , and so on . Nonetheless , we still have much to learn about the information processes of existing CNNs and how they relate to cortical sensory functions , and there is fruitful work to be done in developing techniques that leverage these models for theoretical insights . Another limitation of this work is that many computer-vision models , and most visual neuroscience experiments , are restricted to simple perceptual tasks using static images . This ignores many important aspects of natural vision that any comprehensive computational model will ultimately need to account for , including attention , motion , temporal dependencies , and the role of memory . Our findings are also limited by the noise ceiling of our neural data . Although the CNN explained as much variance in the OPA as could be expected for any model , there still remains a large portion of variance that can be attributed to noise . This noise arises from multiple factors , including inter-subject variability , variability in cortical responses across stimulus repetitions , the limited resolution of fMRI data , signal contamination from experimental instruments or physiological processes , and irreducible stochasticity in neural activity . Improving the noise ceiling in fMRI studies of high-level visual cortex will be an important goal for future work . Several of these noise-related factors could potentially be mitigated through the use of larger and more naturalistic stimulus sets or through improved data pre-processing procedures . Others , such as inter-subject and inter-trial variability may have identifiable underlying causes that are important for understanding the functional algorithms of high-level visual cortex [55 , 56] . In addition to the experimental approaches used here , there are other important avenues of investigation for relating the computations of neural networks to the visual system . For example , CNNs could be used to generate new stimuli that are optimized for testing specific computational hypotheses [17] , and then fMRI data could be collected to examine the role of these computations in visual cortex . Another useful approach would be to run in silico lesion studies on CNNs to understand the role of specific units within a computational system . Finally , an important direction for future work will be to use the conclusions of experiments on CNNs to build simpler models that embody specific computational principles and allow for detailed investigations of the necessary and sufficient components of information processing in vision . An important goal of neuroscience is to understand the computational operations of biological vision . In this work , we utilized recent advances in computer vision to identify an image-computable , quantitative model of navigational-affordance coding in scene-selective visual cortex . By running experiments on this computational model , we characterized the stimulus inputs that drive its internal representations , and we revealed the complex , high-level scene features that its computations give rise to . Together , this work suggests a computational mechanism through which visual cortex might encode the spatial structure of the local navigational environment , and it demonstrates a set of broadly applicable techniques that can be used to relate the internal operations of deep neural networks with the computational processes of the brain .
We analyzed human fMRI data that were described in a previous publication , which includes a complete description of IRB approval and subject consent [19] . We used RSA to characterize the navigational-affordance information contained in multivoxel fMRI-activation patterns and multiunit CNN-activity patterns . For the fMRI data , we extracted activation patterns from a set of functionally defined ROIs for each of the 50 images in the stimulus set , using the procedures described in our previous report [19] . Briefly , 16 subjects viewed 50 images of indoor scenes presented for 1 . 5 s each in 10 scan runs , and performed a category-detection task . Subjects were asked to fixate on a central cross that remained on the screen at all times and press a button if the scene they were viewing was a bathroom . A general linear model was used to extract voxelwise responses to each image in each scan run . ROIs were based on a standard set of functional localizers collected in separate scan runs , and they were defined using an automated procedure with group-based anatomical constraints [19 , 57 , 58] . For each subject , the responses of each voxel in an ROI were z-scored across images within each run and then averaged across runs . We then applied a second normalization procedure in which the response patterns for each image were z-scored across voxels . Subject-level RDMs were created by calculating the squared Euclidean distances between these normalized response patterns for all pairwise comparisons of images . The squared Euclidean distance metric was used ( here and for the other RDMs described below ) because several of our analyses involved multiple linear regression for assessing representational similarity . In this framework , the distances from one RDM are modeled as linear combinations of the distances from a set of predictor RDMs . This requires the use of a distance metric that sums linearly [6 , 59] . Squared Euclidean distances sum linearly according to the Pythagorean theorem , and when the representational patterns are normalized ( i . e . , z-scored across units ) , these distances are linearly proportional to Pearson correlation distances , which we used in our previous analyses of these data [19] . We then constructed group-level neural RDMs for each ROI by taking the mean across all subject-level RDMs . The use of group-level RDMs allowed us to apply the same statistical procedures for assessing all comparisons of RDMs ( i . e . , fMRI RDMs , navigational-affordance RDM , and CNN RDMs ) . Furthermore , the use of group-level RDMs , which are averaged across subjects , has the benefit of increasing signal-to-noise and improving model fits for the RSA comparisons . To construct RDMs for each layer of the CNN , we first ran the experimental stimuli through a pre-trained CNN that can be downloaded here: http://places . csail . mit . edu/model/placesCNN_upgraded . tar . gz . We recorded the activations from the final outputs of all linear-nonlinear operations within each layer of the CNN . All layers , with the exception of layer 8 , contain thousands of units . We found that the RSA correlations between the layers of the CNN and the ROIs were improved when the dimensionality of the CNN representations was reduced through principal component analysis ( PCA ) . This likely reflects the fact that all CNN units were weighted equally in our calculations of representational distances , even though many of the units had low variance across our stimuli . PCA reduces the number of representational dimensions and focuses on the components of the data that account for the largest variance . We therefore set the dimensionality of the CNN representations to 45 principal components ( PCs ) for each layer . Our findings were not contingent on the specific number of PCs retained; we observed similar results across the range from 30 to 49 PCs . We z-scored the CNN activations across PCs for each image and calculated squared Euclidean distances for all pairwise comparisons of images . The neural and CNN RDMs were compared with an RDM constructed from the representations of a navigational-affordance model . To construct this model , we calculated representational patterns that reflected the navigability of each scene along a set of angles radiating from the bottom center of the image ( Fig 1 ) . These navigability data were obtained in a norming study in which an independent group of raters , who did not participate in the fMRI experiment , indicated the paths that they would take to walk through each scene ( Fig 1B ) [19] . In our previous report , we combined these navigational data with a set of idealized tuning curves that reduced the dimensionality of the data to a small set of hypothesized encoding channels ( i . e . , paths to the left , center , and right ) . Here , however , we used a different approach in which we simply smoothed the navigability data over the 180 degrees of angular bins using an automated and robust smoothing method [60] . This smoothing procedure was implemented using publicly available software from the MATLAB file exchange: https://www . mathworks . com/matlabcentral/fileexchange/25634-fast—n-easy-smoothing ? focused=6600598&tab=function . We then z-scored these smoothed data across the angular bins for each image and calculated squared Euclidean distances for all pairwise comparisons of images . For standard RSA comparisons of two RDMs , we calculated representational similarity using Spearman correlations . The Spearman-correlation procedure assesses whether two models exhibit similar rank orders of representational distances , which allows for the possibility of a nonlinear relationship between the pairwise distances of two RDMs . Nonetheless , we observed similar results using Pearson correlations or linear regressions , and thus our RSA findings were not contingent on the use of a non-parametric statistical test . Bootstrap standard errors of these correlations were calculated over 5000 iterations in which the rows and columns of the RDMs were randomly resampled . This is effectively a resampling of the stimulus labels in the RDM . Resampling was performed without replacement by subsampling 90% of the rows and columns of the RDMs . We did not use resampling with replacement because it would involve elements of the RDM diagonals ( i . e . , comparisons of stimuli to themselves ) that were not used when calculating the RSA correlations [61] . All bootstrap resampling procedures were performed in this manner . Statistical significance was assessed through a permutation test in which the rows and columns of one of the RDMs were randomly permuted and a correlation coefficient was calculated over 5000 iterations . P-values were calculated from this permutation distribution for a one-tailed test using the following formula: pvalue=∑ ( Rperm≥Rtest ) +1N+1 where Rperm refers to the correlation coefficients from the permutation distribution and Rtest refers to the correlation coefficient for the original data . All p-values were Bonferroni-corrected for the number of comparisons performed ( i . e . , the number of ROIs in Fig 1 and the number of CNN layers in Fig 2 ) . We calculated the noise ceiling for RSA correlations in the OPA as the mean correlation of each subject-level OPA RDM to the overall group-level OPA RDM , a measure that reflects the inherent noise in the fMRI data [62] . According to this metric , the best-fitting model for an ROI should explain as much variance as the average subject . Several analyses involved the use of multiple linear regression and a variance-partitioning procedure to quantify the overlap of explained variance for two predictor RDMs . The multiple linear regression models included two regressors for the predictor RDMs and a third regressor for the constant term . These regressors were used to explain variance in a third RDM , which served as the dependent variable . Thus , the data points for the dependent and independent variables were the pairwise distance measurements of the RDMs . The models were fit using ordinary least squares regression . We quantified the overlap of explained variance for the two predictor RDMs using a procedure known as commonality analysis [32] . This procedure partitions the explained variance of the regression model into the shared and unique components contributed by all regressors . We used this analysis to determine the degree to which the explained variance of one regressor ( e . g . , the affordance RDM ) was shared with a second regressor ( e . g . , the CNN RDM ) . We refer to this quantity as shared variance , and we calculated it using the following formula: SV=100*γ12γ12+γ1 where SV is the percentage of the explained variance for regressor X1 that is in common with regressor X2 ( see also Fig 3A ) . The other variables in this equation refer to components of the overall explained variance ( R212 ) : γ1 = unique contribution of X1 to R212 γ12 = common contribution of X1 and X2 to R212 These values are calculated as follows: γ1=R122−R22 γ12=R12+R22−R122 where R212 is the explained variance of a regression model with both X1 and X2 , R21 is the explained variance of a model with only X1 , and R22 is the explained variance of a model with only X2 . Bootstrap standard errors of this shared-variance metric were calculated over 5000 iterations in which the rows and columns of the RDMs were randomly resampled and the variance-partitioning procedure was applied to the resampled RDMs . We quantified the contribution of specific low-level image features to the RSA effects of the CNN . To do this , we generated new sets of filtered stimuli in which specific visual features or portions of the image were isolated or removed ( e . g . , color , spatial frequencies , edges at cardinal or oblique orientations , lower or upper portions of the image; Figs 4 and 5 ) . These filtered stimuli were passed through the CNN , and new RDMs were created for each layer . We used the commonality-analysis technique described above to quantify the portion of the original explained variance of the CNN that could be accounted for by the filtered stimuli . This procedure was applied to the explained variance of the CNN for predicting both the navigational-affordance RDM and the OPA RDM . We performed five different stimulus transformations to examine specific classes of image features . The first was a simple transformation of the images from color to grayscale that allowed us to assess the importance of color information . The others reflect two broad categories of low-level image properties: spatial frequencies and contour orientations . To examine the role of spatial frequencies , we created one set of stimuli in which low spatial frequencies were removed from the images ( high-pass ) and another set in which high spatial frequencies were removed ( low-pass ) . These were created by first converting the images to grayscale , performing a Fourier transform , filtering out a subset of frequencies , and then reconstructing the grayscale images from the filtered Fourier transforms . For the high-pass images , the Fourier spectrum was filtered using a Gaussian filter with a standard deviation set at 0 . 1 cycles per pixel . A similar approach was used for the low-pass images , with the standard deviation of the Gaussian filter set at 0 . 0075 cycles per pixel . To examine the role of contour orientations , we created one set of filtered stimuli in which edges at cardinal orientations were emphasized ( cardinal ) and another set in which edges at oblique orientations were emphasized ( oblique ) . These were created by first converting the images to grayscale and then performing a convolution to extract image contours at cardinal orientations ( 0 and 90 degrees ) or oblique orientations ( 45 and 135 degrees ) . The convolution kernels spanned 3 pixels by 3 pixels and are depicted in S8 Fig . Convolutions were performed separately for the two orientations in each set ( e . g . , 0 and 90 degrees ) and a combined output was created by squaring and summing these convolutions and then taking the square root of their sum . We statistically assessed differences in shared variance across sets of filtered images by calculating confidence intervals on their difference scores . We did this for the following subsets: 1 ) high-pass minus low-pass and 2 ) cardinal minus oblique . To do so , we calculated a bootstrap distribution of the difference in shared variance values across each image set over 5000 iterations in which the rows and columns of the RDMs were randomly resampled . From this distribution , we computed the value of the lower 95th percentile for a one-tailed test to determine if the 95% confidence interval was above zero . We also performed analyses to examine the importance of visual inputs at different positions along the vertical axis of the image . To do this , we generated occluded versions of the stimuli in which everything outside of a small horizontal slice of the image was masked . The exposed slice of the image spanned 41 pixels in height , which was 18% of the overall image height . We used commonality analysis to quantify the portion of the original explained variance of the CNN that could be accounted for by the occluded stimuli . This procedure was repeated with the un-occluded region shifted by a stride of 5 pixels on each iteration until the entire vertical axis of the image was sampled . We generated heat maps of these results by assigning shared variance values to the pixels in each horizontal slice and then averaging the values across overlapping slices . We used a receptive-field mapping procedure in combination with a set of data-driven visualization techniques to gain insights into the complex feature selectivity of units within the CNN . The receptive-field mapping and image-segmentation procedures were based on previously published methods [42] . We mapped the selectivity of individual CNN units across each image by iteratively occluding the inputs to the CNN . First , the original image was passed through the CNN . Then a small portion of the image was occluded with a patch of random pixel values of size 11 pixels by 11 pixels , as in [42] . The occluded image was passed though the CNN , and discrepancies in unit activations relative to the original image were logged . Theses discrepancies were calculated as the absolute value of the difference in activation , which is consistent with the procedure used by Zhou and colleagues [42] ( personal communication with Bolei Zhou ) . On each iteration , the position of the occluding patch was shifted by a stride of 3 pixels . After iteratively applying this procedure across all spatial positions in the image , a two-dimensional discrepancy map was generated for each unit and each image ( Fig 6A ) . Each discrepancy map indicates the sensitivity of a CNN unit to the visual information across all spatial positions of an image . The spatial distribution of the discrepancy effects reflects the position and extent of a unit’s receptive field , and the magnitude of the discrepancy effects reflects the sensitivity of a unit to the underlying image features . We generated image segmentations to visualize the scene features that individual CNN units were most sensitive to ( Fig 6B ) . We first smoothed the discrepancy maps by convolving them with a local averaging filter of 20 pixels by 20 pixels . For each unit , we then selected the 3 stimulus images that generated the largest discrepancy values at any spatial location in the image . We segmented these discrepancy maps by identifying pixels with a discrepancy value equal to at least 10% of the peak discrepancy across all pixels . We generated these visualizations for 50 units in layer 5 . These units were chosen based on their unit-wise RSA correlations to the affordance RDM and the OPA RDM ( we chose the units with the highest mean correlation to these two RDMs ) . We then used t-SNE and k-means clustering to generate a summary visualization and to identify common themes among the scene features that were highlighted by these segmentations [63] . Our goal was to cluster the image segmentations based on the similarity of their high-level scene content . We first created image patches of 81 pixels by 81 pixels centered on the peak discrepancy value for the top 3 images for each unit . We ran these image patches through the CNN and logged the responses in layer 5 . These responses were then averaged across the top 3 images for each unit , and t-SNE was used to generate a two-dimensional embedding of all 50 units based on the similarity of their mean response vectors from layer 5 . We assigned the units in this embedding to clusters with similar scene features ( Figs 6B and S1–S7 ) . Clusters were identified in a data-driven manner through k-means clustering , with the number of clusters chosen ( within the range of 1 to 10 clusters ) using the silhouette criterion in the MATLAB function evalclusters . We sought to determine whether the affordance-related components of the CNN identified in our previous analyses of indoor scenes could also be used to identify the navigational properties in a set of highly varied outdoor environments . To do this , we examined the classification of navigability in natural landscapes using a stimulus set from a previous study [43] . These stimuli are described in detail elsewhere [43 , 64] . Here we summarize their key properties . All stimuli depict outdoor environments , and they were sampled from a diverse set of semantic categories . They were selected based on behavioral data to represent the poles ( i . e . , low and high ) of 14 high-level scene properties , which included global properties , such as navigability , and semantic categories , such as forest . To create these stimulus groups , a large set of images were ranked according to their representativeness for each scene property , based on the subjective assessments of human observers . Images were selected from the lower and upper ranks of representativeness to create binary classes of “low” and “high” ( n = 50 images in each class ) . There were 548 images in total , and many were used across multiple scene properties . We examined how well these scene properties could be classified from the responses of the CNN . Our analyses focused on the responses in layer 5 , which was the layer that showed the strongest effects for navigational-affordance representations in our previous analyses . To begin , we evaluated the classification accuracies obtained when using the responses of all units in layer 5 . For this analysis , we used a simple minimum-distance classifier , which avoided the challenges of fitting classifier parameters to a data set in which the number of feature dimensions is much larger than the number of stimuli . All images were run through the CNN , and the activations of layer 5 were recorded . For each scene property , we constructed an RDM for its 100 associated images . The unit activations were first z-scored across images and then z-scored across units . RDMs were created by calculating the pairwise Euclidean distances between images based on these unit activations . For each scene property , images were classified into categories of “low” or “high” based on their mean pairwise distance to all other images in those categories . For example , to classify an image in the navigation set , we calculated its mean pairwise distances to all low-navigability images and to all high-navigability images ( excluding the comparison of the image to itself ) . The image was then classified into the category with the smallest mean pairwise distance . Using this approach , navigability was classified at an accuracy of 86% , and all other scene properties were classified at an accuracy of 86% or higher . Thus , the representations of layer 5 appear to be informative for a broad range of high-level scene properties , including the navigability of natural landscapes . We next examined a small subset of units in layer 5 that we found to be particularly informative in our previous analyses . Specifically , we examined the 50 units that were previously selected for the visualization analyses in Fig 6 . For each scene property , we quantified the classification accuracy obtained when passing the responses of these units to a linear discriminant analysis in a leave-one-out cross-validation design . In this procedure , a single image was held out and a linear discriminant classifier was fit to the other 99 images . This classifier was then used to generate a prediction for the held-out image ( i . e . , low or high ) , and the procedure was repeated for each image in turn . Accuracy was calculated as the percentage of correct classifications across all images . For comparison , we also computed the classification accuracies obtained from randomly selected subsets of units . Specifically , we performed an iterative resampling procedure in which 50 units were randomly selected from layer 5 ( without excluding our original 50 units ) and then submitted to the same leave-one-out classification analyses described above . This procedure was repeated 5 , 000 times for each scene property . The resulting accuracy distributions were visualized using kernel density estimation , as implemented in the MATLAB function ksdensity ( Fig 7A ) . For the classification of navigability , we observed only a single instance of a resampled accuracy score that exceeded the performance of our original 50 units . Thus , the classifier performance of the affordance-related units was in the 99th percentile of this resampling distribution . For all other scene properties , the classification accuracy of the affordance-related units was no higher than the 69th percentile . | How does visual cortex compute behaviorally relevant properties of the local environment from sensory inputs ? For decades , computational models have been able to explain only the earliest stages of biological vision , but recent advances in deep neural networks have yielded a breakthrough in the modeling of high-level visual cortex . However , these models are not explicitly designed for testing neurobiological theories , and , like the brain itself , their internal operations remain poorly understood . We examined a deep neural network for insights into the cortical representation of navigational affordances in visual scenes . In doing so , we developed a set of high-throughput techniques and statistical tools that are broadly useful for relating the internal operations of neural networks with the information processes of the brain . Our findings demonstrate that a deep neural network with purely feedforward computations can account for the processing of navigational layout in high-level visual cortex . We next performed a series of experiments and visualization analyses on this neural network . These analyses characterized a set of stimulus input features that may be critical for computing navigationally related cortical representations , and they identified a set of high-level , complex scene features that may serve as a basis set for the cortical coding of navigational layout . These findings suggest a computational mechanism through which high-level visual cortex might encode the spatial structure of the local navigational environment , and they demonstrate an experimental approach for leveraging the power of deep neural networks to understand the visual computations of the brain . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"diagnostic",
"radiology",
"functional",
"magnetic",
"resonance",
"imaging",
"neural",
"networks",
"engineering",
"and",
"technology",
"applied",
"mathematics",
"brain",
"social",
"sciences",
"neuroscience",
"magnetic",
"resonance"... | 2018 | Computational mechanisms underlying cortical responses to the affordance properties of visual scenes |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.